GILLETTE, J.
In this termination of parental rights proceeding, the issue is whether the state has shown, by clear and convincing evidence, that, at the time of the termination proceeding, mother was unfit and unable to be a minimally adequate parent to her daughter. On de novo review, the Court of Appeals concluded that mother was unfit because she has a mental illness and an addiction to controlled substances that, together, are seriously detrimental to her child. The Court of Appeals therefore affirmed the trial court’s order terminating mother’s parental rights. State ex rel Dept. of Human Services v. Simmons, 203 Or App 279, 125 P3d 66 (2005) (Simmons III).1 For the reasons that follow, we reverse the decision of the Court of Appeals.
In reviewing a decision of the Court of Appeals in a parental rights termination case, this court may review de novo or it may limit its review to issues of law. ORS 19.415(4); State ex rel SOSCF v. Stillman, 333 Or 135, 138, 36 P3d 490 (2001). Because we conclude that neither the trial court nor the Court of Appeals correctly considered the evidence in the record concerning mother’s condition and its effect on child at the time of the termination trial, we elect to view the record de novo.
At the time of the trial in this case, mother was 38 years old. Since the age of nine, mother has suffered from a chronic, painful, autoimmune disease. In addition, in adulthood, mother suffered other debilitating illnesses, including pancreatitis, hepatitis C, and sleep apnea. Because of her health conditions, mother had not worked and had been receiving disability benefits since 1992. Over the years, various doctors prescribed pain management medications for [79]*79mother, including the opiate OxyContin, which is an oral, time-release form of the drug Oxycodone.
In around 1985, mother became involved with Keith Woodman, child’s father. Woodman was addicted to heroin and physically abused mother. In 1995, mother gave birth to child. Four months later, she left Woodman. Woodman never paid child support and has had little contact with child.2
Because of mother’s health problems, she required the assistance of a caregiver after leaving Woodman. Mother and child moved in with mother’s sister, Cynthia Barton, in Everett, Washington, so that Barton could help mother with her daily needs. By June 2000, Barton no longer was able to care for mother, and mother and child moved to an apartment in Baker City, Oregon, near the home of mother’s other sister, Crystal Dahlem. Dahlem then became mother’s paid caregiver. She provided daily assistance, dispensing mother’s medications and cooking and cleaning for her. Mother’s and Dahlem’s relationship quickly began to deteriorate, however, and they argued loudly and frequently, often in front of child. It also became apparent during that time that mother was abusing her pain medications.
Soon thereafter, mother and child moved into the home of another woman whom mother paid to help her with her daily household needs. In December 2000, that person called the police to report mother’s drug use. Police responded and found evidence that mother had been crushing OxyContin tablets to defeat the time-release mechanism, then “cooking” the resulting powder to a liquid, loading it into a syringe, and then injecting herself with it. Mother also appeared to be using OxyContin at a rate far exceeding the prescribed dosage. The police called DHS and DHS temporarily placed child with Dahlem.
Mother then entered an inpatient drug treatment program, taking child with her. Mother was unsuccessful in that program. Among other things, mother would not admit that she had a drug problem; on her intake assessment form, mother reported that she had last abused OxyContin four years earlier. In addition, counselors observed that mother [80]*80inappropriately relied on child, then six years old, to assume the role of caretaker. After three weeks, mother was discharged from the program for noncompliance with program requirements, for denying her addiction, and for engaging in drug-seeking behaviors.
Mother and child returned home to mother’s apartment and mother hired yet another paid caregiver, Lorrie Moss, to assist her. Moss dispensed mother’s medications, cleaned her house, did the laundry, did the grocery shopping, and took mother to appointments. She worked for mother eight hours a day, five to seven days per week. During that time, mother was extremely sleepy and occasionally felt too unwell even to get out of bed. On those days, Moss picked child up from school. From time to time, mother failed to pick child up from school without making other arrangements and Moss would receive an urgent call from the school.
In June 2001, Moss found syringes wrapped in paper towels in a drawer, although mother had not been prescribed any medicines that required mother to use syringes. Moss disposed of them in a dumpster outside mother’s house, but the next day they were back in the bathroom; mother had retrieved them from the garbage. Mother then repeatedly lied, to Moss and others, about how she came to possess the needles. Around that time, child, who then was six years old, saw mother inject herself with drugs and, on more than one occasion, found mother asleep on the sofa and was unable to rouse her. Child became frightened that mother would die. She began to sleep over at a neighbor’s house. Moss called DHS to report mother’s drug abuse. Moss also reported that she had witnessed volatile arguments between mother and Dahlem over money and care of child. DHS removed child at the end of June 2001, placing her in a neutral foster home in Baker City.
Mother then began another outpatient drug treatment program, again with little success. Mother continued to deny her addiction and missed appointments and group meetings. She still was abusing OxyContin during the months after child was removed from mother’s home and, [81]*81during that time, mother was arrested for possessing a small bindle of methamphetamine.3
In August 2001, mother met with Dr. David Sweet, a psychologist whom the state had hired to evaluate her. During the interview, mother was trembling and had difficulty focusing on and answering Sweet’s questions. Mother minimized her abuse of OxyContin and denied any other drug use or abuse. Sweet administered various tests, which mother took an extraordinarily long time to complete. The results of mother’s personality test were of dubious validity but did show mother to be extremely defensive and obsessive, and desirous of presenting herself in a positive light. Mother admitted to Sweet that she was depressed. According to Sweet, mother’s pattern of moving from place to place and changing caregivers showed her inability to provide stability for herself or child. Mother appeared to Sweet to be overwhelmed by her problems and to lack insight into them, making it likely that she would continue to make poor choices. Sweet diagnosed mother with polysubstance dependence and major depression, and made what he called a “rule-out” diagnosis of borderline or dependent personality disorder.4 Sweet recommended that child remain in foster care until mother could show for an extended period of time that she could provide stability for child and address her substance abuse problem.
A few days later, Sweet also evaluated child. He based his evaluation on background information DHS had provided him, on his own observations and testing of child, and on an interview with child’s foster mother. The foster mother reported that child seemed well adjusted, all things considered; she interacted well with others and was settling in with her peer group. However, child wet the bed at night and wet herself during the day. Sweet’s interview with child herself revealed that child frequently had seen mother inject herself with drugs, that mother slept a lot and often was [82]*82difficult to wake up, which frightened child, and that they moved around a lot. However, child reported that she was happy “most of the time,” and she stated that living with mother was “fun, great, good.”
The testing showed that child scored in the average range on most tests, but on three verbal subtests and on certain nonverbal tests, child scored in the low-average range. According to Sweet, a test requiring child to tell stories about a series of ambiguous pictures revealed that child was sad, angry and depressed, and that she felt vulnerable and worried about her living arrangements. Child also appeared to be closed and inaccessible, finding it difficult to discuss personal information.
Based on the foregoing, Sweet concluded that child was very bright but that she had had serious problems resulting from mother’s condition, which interfered with her ability to learn and gather information about her environment. He agreed with a social worker’s impression that child was “par-entified,” meaning that child believed that she had to take care of her mother. According to Sweet, it would be best for child to remain in foster care until mother could provide a stable home in which mother clearly was the parent and child clearly was the child.
In September 2001, child began therapy with a social worker, Karen Wain. In play and in stories that she told, child conveyed to Wain that child was afraid of being abandoned. She also told Wain that she often had heard mother arguing with her sister, Dahlem, and that she was afraid of Dahlem. At that time, child occasionally had unsupervised visits with Dahlem and Dahlem’s daughter, and her bed and daytime wetting worsened after those visits.
Child also began to have supervised visits with mother at DHS offices in Baker City. The record contains the observations of three DHS workers during that period. All reported seeing a consistently strong, loving, and affectionate relationship between mother and child. However, one observer noted that mother occasionally talked to child in detail beyond what a child of her age could comprehend and that mother once teased child inappropriately about marrying a boy in her class one day. The second observer, a DHS [83]*83social worker, stated that, while mother would cook nutritious and appropriate dinners for child dining the visits, on one occasion, she became distracted by cooking and did not pay the child close attention. That social worker also noted that mother occasionally talked to child about child’s boyfriends, even though child was only six years old at the time, and that mother let child dress up in scanty attire and dance to popular music.
Finally, the third observer, also a social worker who supervised several visits at mother’s home in Baker City during that period, found mother’s behavior during the visits to be entirely appropriate. However, in December 2001, that social worker observed one incident that raised a safety concern for her: she noticed a syringe on the top of the medicine cabinet in the bathroom. She did not say anything at the time, but she called authorities and, two days later, the police searched mother’s house. Police found hypodermic needles in the bathroom and two bindles of methamphetamine between the mattress and the box spring of mother’s bed. Mother denied that the drugs were hers. Mother was arrested but the charges against her later were dismissed.
In late December 2001, DHS moved child to a new foster home in Lake Oswego: that of child’s paternal aunt (Woodman’s sister) and the aunt’s husband, the Stubblefields. Mother initially agreed to that placement because she had hoped to find a drug treatment center in Portland that could address her medical problems in addition to her substance abuse problem. She was unable to find such a program, however, and she remained in Baker City.
After child moved to Lake Oswego, DHS arranged for her and the Stubblefields to undergo family counseling with Dr. Christine Portland, a psychologist. The counseling focused mostly on child, but Portland also met with the Stubblefields to address issues related to the transition. Portland diagnosed child with post-traumatic stress disorder, based on child’s reports of having witnessed and been frightened by mother’s drug use and loud arguments between mother and Dahlem. She also noted child’s enuresis. In addition, Portland found child to be delayed in her emotional development; Portland concluded that child had a [84]*84chronologic age of seven but an emotional age at least a year- and-a-half younger. Accordingly, Portland at first worked with the Stubblefields to ensure that child felt safe in her new home, and then focused her attention on helping child in her social and emotional development.
Based on her therapy sessions with child, Portland concluded that child was anxious about her mother, about Dahlem, and about Dahlem’s daughter, as well as about the structure of her daily life. Portland concurred that the concern that child demonstrated for her mother and others was an expression of her parentification and that child’s parenti-fication caused her pain, manifested as irritability, defiance, and bed-wetting.
After child moved to Lake Oswego, mother became depressed and, by February 2002, was suicidal. At that point, with the help of her sister and her lawyer, mother entered an inpatient psychiatric hospital program in Boise, Idaho. There, mother participated in drug treatment and, at the same time, doctors addressed mother’s ongoing medical conditions in light of her addiction. Doctors adjusted mother’s medications and mother began to improve significantly both mentally and physically. That improvement continued to the time of the trial.
Immediately after mother’s successful completion of and release from that program, mother began participating in outpatient drug treatment again in Baker City, beginning in March 2002. This time, counselors reported that mother was compliant with program rules, attended virtually all meetings, and followed instructions. She underwent regular urinalysis tests; all were negative. Mother obtained a sponsor and completed a 12-step program. Mother continued the treatment through September 2002, when she graduated from the program. On the discharge sheet, mother’s counselor wrote that all problems had been addressed.
During the time after mother’s release from the Boise inpatient program, mother and child had more than 50 supervised telephone conversations as well as supervised visits, either at a hotel in Clackamas or at DHS offices in Baker City. The DHS social worker who listened in on the telephone calls reported that both mother and child seemed to enjoy the [85]*85calls, that they were affectionate and loving toward each other, that mother planned for the telephone calls by having books available to read to child, and that mother complied with DHS rules for her behavior during the calls.
The supervised visits lasted five hours on Saturdays and two hours on Sunday morning. They were scheduled irregularly at first but eventually they took place every other weekend. Beginning in spring 2002, DHS social worker Wain, who had provided therapy for child immediately after she was removed from mother’s care, supervised the visits in Baker City. Wain observed that mother and child were loving together, that mother brought age-appropriate activities to the visits, that mother prepared nutritious lunches, and that she was able to play with child appropriately. Wain nevertheless found mother’s parenting skills lacking, because Wain did not think that mother was able to distinguish between her own and child’s emotional needs. As evidence of that fact, Wain referred to an instance when mother and child were playing with sidewalk chalk and mother disregarded child’s express wish that mother not write on the sidewalk that she loved child. In addition, Wain pointed to child’s “passive aggressive” actions when she and mother played, and discounted the affection that child demonstrated toward mother as “placating” behavior stemming from her parentification.
From May through October 2002, DHS contracted the supervision of the Clackamas visits to an outside agency, Options Counseling Services of Oregon, Inc. (Options). From May through August 2002, Options counselor Velinda Sloan supervised seven visits. Sloan observed that mother and child were extremely affectionate with each other, and that that affection seemed to her to be genuine. She reported that mother played well with child, engaged child in appropriate discussions about child’s life, disciplined child appropriately as necessary, and refrained from talking about her own problems. Sloan did not report seeing anything that alarmed her in mother’s parenting.
From August through October 2002, Options family resource worker Veronica Roman supervised the visits. Her observations were consistent with Sloan’s. She reported one negative incident, in which Roman arrived for the visit with [86]*86child and mother was asleep in her hotel room and was difficult to rouse. Mother apologized and explained that she must have set her alarm incorrectly, and the matter was resolved.
In July 2002, DHS arranged for Sweet to examine both child and mother again. Sweet first met with mother, interviewing her and subjecting her to a battery of tests. Sweet observed that mother was cooperative and responded more appropriately than she had the last time that he had seen her. Mother told Sweet that she was feeling much stronger than she had felt before, that she was managing her own home, and that she was riding her bicycle all over town. When asked to explain the problems that had led child to be taken away from her, she admitted that she had been “shooting up” her OxyContin. She stated that she had stopped after child was removed, but admitted that she had relapsed at some point thereafter. At the time of the interview, mother had not abused any of her medications in about six months. Mother also reported that DHS’s current concern was with her ability to meet child’s physical and emotional needs. Finally, mother described her current relationship with Dahlem as greatly improved. She explained that she and Dahlem had learned to talk to each other and to trust each other, and mother was proud of the fact that Dahlem permitted mother to babysit her own daughter in either of the women’s homes. When Sweet asked mother to what she attributed that improvement, mother responded that, when she had been abusing drugs, she would pick fights with Dahlem, but now she realized that Dahlem had only been trying to help her and now they were both making an effort to get along.
Mother’s various psychological test results in most instances were similar to those of the year before. However, mother now was able to complete the personality test in a reasonable amount of time. She continued to provide a very defensive profile, although not as defensive as the year before. Still, the current results showed that mother was a very rigid, over-controlled woman, and, in people like that, according to Sweet, one sees a tendency to overreact aggressively or even violently. Based on those test results and on background information provided him by Wain, Sweet diagnosed mother with a “personality disorder, n[ot] otherwise] [87]*87specified], with characteristics of borderline, dependent and narcissistic personality disorders,” that interferes with her ability to function and to get along with others.
A few days later, Sweet met with child. He observed that child was verbal and cooperative during the session. Sweet administered a series of tests that showed that child is in the high-average intelligence range and showed strength on verbal skills tests. Other tests revealed that child still was somewhat depressed and worried about her situation but that she had improved in that respect over the previous year. Those tests also showed that child felt hopeful about the future and that she felt good about herself.
In around August 2002, DHS decided to change its plan for mother and child from reunification to achieving adoption of child. Sandi Baer, the DHS supervisor in Baker City, stated that DHS based that decision on the following factors: (1) a May 2002 statement by an addiction counselor that mother “wasn’t buying into the disease concept” and therefore he did not believe that child safely could be returned to her;5 and (2) Sweet’s second evaluation of mother in July 2002. Baer also referred to the facts that mother did not begin bringing lunches to visitations until July 2002 despite the agency’s repeated demands that she do so and that mother did not think that her parenting sessions with Wain were profitable because she already knew the things that Wain was telling her.
Once DHS had made the decision to seek termination of mother’s parental rights, but before ever meeting mother for the first time, Portland recommended that telephone visits between mother and child cease completely and that supervised visitations be cut back from seven hours every other weekend to one and a half hours every other month.
Portland supervised the first visit after the cutback, as well as two subsequent visits. Mother brought presents for child to those visits and played with her for the entirety of the [88]*88visits. Portland testified at trial that that behavior on mother’s part was evidence of her inability adequately to parent child, insofar as mother brought only “fun” and not practical presents and that mother used the time to play with child rather than to explore child’s feelings.
In April 2003, the state petitioned to terminate mother’s parental rights under ORS 419B.504. The state alleged that mother is unfit due to (1) addictive or habitual use of controlled substances to the extent that her parenting ability is substantially impaired; (2) lack of effort or failure to maintain a suitable or stable living situation; (3) failure to present a viable plan for return of the child to her care; (4) an emotional or mental illness or deficiency of such nature and duration as to render her incapable of providing care of the child for an extended period of time; (5) physical and emotional neglect of child; (6) lack of effort to adjust circumstances, conduct, or conditions to make return of the child possible; (7) failure to effect a lasting adjustment after reasonable efforts by available social services have been made; and (8) multiple medical needs that, coupled with mother’s mental health and drug abuse problems, substantially impaired mother’s ability to care for the child.
The case went to trial in October 2003. A large part of the state’s case was devoted to establishing the severity of mother’s drug abuse problem before child was removed from her care in June 2001 (and the effect of that problem on child) and mother’s early failures in acknowledging and addressing that problem. In addition, the state presented several witnesses who testified about mother’s mental state, her behavior during visits with child, and her compliance with DHS service agreements. The state’s case began with testimony from Wain, who had last seen mother in October 2002, a year before the trial. Wain testified that mother did not have the ability to separate her own emotional needs from those of child as evidenced by, among other things, the fact that mother sometimes broke the agency’s rule against talking about her “adult decisions” with child and that she failed to acknowledge that child was afraid of Dahlem. Wain also reported that child seemed not to talk about her own emotions with mother and inferred from that fact that child did not feel safe expressing those feelings to mother. Wain [89]*89concluded that, in her opinion, mother did not have the parenting skills necessary to be an adequate parent and that returning child to her care would be detrimental to child’s welfare.
The psychologist, Sweet, also testified for DHS. Like Wain, Sweet had not seen mother in some time. Based on his last evaluation of mother in July 2002 (the last time Sweet saw mother), Sweet testified that mother suffers from a personality disorder that would make it difficult for mother to get along with others or to place child’s needs above her own, and that personality disorders are resistent to treatment. Based on information about mother that Sweet had been provided before his evaluation, Sweet concluded that mother “didn’t have a clue” about how her behavior had affected child.
Portland, child’s therapist, testified at trial about her early impressions of child and those that she had formed from three meetings with child in the weeks immediately before the trial. Portland observed that child had improved academically in her time in foster care and that her enuresis was under control. Portland also saw improvement in child’s emotional development: child seemed to be happy, there was a sparkle in her eyes, she had friends, and she was able to talk about her feelings, both negative and positive. Portland reported that child was beginning to be inattentive and disorganized at school, but admitted that that behavior likely was due to uncertainty brought about by the impending termination proceeding.
Portland also testified at length about why she believed that child needed to be protected from mother. She stated that mother was controlling and rigid and not able to empathize with child’s feelings or put child’s needs over her own, as evidenced by the following facts: (1) child wanted to have her hair cut but mother wanted child to keep it long and believed that parents have the right to make decisions about their children’s hair; (2) that mother named a puppy that she bought for child “Casey” when child expressed her wish to name it “Max”; and (3) mother made the visits that Portland supervised “fun and games” by bringing presents and playing with child the entire time and did not “come in with herself [90]*90and be with [child], and talk to [child] about who she is now, and talk about the mess that the family’s in, and to be empathetic with [child’s] pain.” Other examples of mother’s inappropriate and insensitive behavior, according to Portland, were that mother once gave child an Easter card that was intended for an intimate adult couple and not for a mother and child, and that she once made an inappropriate joke about “thongs,” when child was talking about sandals. Finally, Portland told a story about an occasion on which child and mother were playing “store” in a dollhouse and child asked mother to go inside. When mother did so, child closed the doors and pulled up a gate. According to Portland, child was showing through her play that she needed to be protected from mother.
Portland reported that, in interactions with mother outside child’s presence, mother did acknowledge all Portland’s concerns about her parenting, but Portland characterized mother’s reaction as “pretty superficial.” Portland also testified that child initially felt responsible for the problems in the family. However, child had had therapy with mother to address that problem and, during that therapy, mother accepted responsibility for all the problems and child now understood that she was not to blame and that mother loved her.
Portland concluded that mother is not capable of providing child with things she needs: a stable home, consistency, structure, and an emotionally mature parent. Portland stated that mother’s deficiencies presently affected child by making her happy and sad in extremes, impairing her ability to trust others, and telling people what she thinks they want to hear. However, the main problem, according to Portland, is the risks to which child will be exposed in the future, during her adolescence, from mother’s poor parenting: she may feel that she is not listened to, she may have poor self-esteem, and she may have a hard time trusting others. And, according to Portland, if child continues to be par-entified, she is at a high risk in adolescence for depression and anxiety, drug use, and even for acquiring HIV and AIDS.
For her part, mother presented evidence of the dramatic improvements that she had made in her life since her [91]*91low point in February 2002. She presented evidence that she had been maintaining a clean, well-kept, and suitable two-bedroom home in Baker City since June 2002. She had completed four stages of drug treatment by September 2002, in which she had not missed a meeting between March and September 2002. Although mother stopped attending AA and NA meetings after she completed her last drug treatment program, mother continued counseling after that until the time of the trial in Baker City with a family therapist, Michael Cross.6 Cross testified that he subjected mother to regular urinalysis tests for illicit substances and that all such tests came back negative, and that he saw mother continue to grow stronger both mentally and physically in the time that he had been seeing her. As of the time of the trial, Cross had no concerns about mother’s ability to parent child.
Mother also presented evidence that, in December 2002, mother began attending twice-weekly services at a local church. Mother eventually assumed leadership responsibilities there and made many friends among the other church members. The senior pastor of that church, Gene Saunders, began to counsel mother weekly. Saunders testified that, in the time that he had known mother, she had matured and was able to communicate appropriately and respectfully and had made and maintained friends. In addition, mother babysat frequently for Saunder’s four children and Saunders never had any concerns for his children’s safety or well-being while they were in mother’s care.
Mother’s family members, including her mother and sister Barton, testified that mother had changed dramatically, both mentally and physically, in the year and a half before the trial. They each testified that mother was more [92]*92insightful, acknowledged her problems, was better able to cope with anxieties, and was much more physically active.
Finally, mother presented the testimony of another psychologist, Dr. Steven Condon, who had interviewed mother twice in August 2003. Condon based his evaluation on Sweet’s earlier evaluations, on Wain’s and Portland’s reports, on DHS’s court report, on a report put together by the Citizen’s Review Board, on his two interviews with mother, and on the results of tests that he had had mother take. Condon diagnosed mother with post-traumatic stress disorder stemming from the physical and emotional abuse mother suffered in her marriage to Woodman. Although Condon observed that mother had a history of major depression, he did not think that she was then suffering from major depression. Condon saw the same results from the personality tests that he had administered as were observed by Sweet, but Condon did not think that there was sufficient information there to enable him to conclude that mother suffers from a personality disorder. (According to Condon, such a diagnosis requires a long-term pattern of inflexibility, criminal history, or severely dysfunctional relationships that was not present here.) Finally, in an offer of proof, Condon testified that, in his professional opinion, he did not believe that there was sufficient evidence on which to conclude that mother lacked the capacity adequately to parent her child.7
In January 2004, the trial court issued a memorandum opinion terminating mother’s parental rights. In its findings of fact, the court first outlined the extensive evidence pertaining to mother’s drug use up through the end of 2001, including the evidence that mother used drugs in front of child. The court also detailed mother’s repeated denials concerning her abuse of OxyContin, including when she last [93]*93had abused that drug, and her continued denials that she ever had used methamphetamine, which persisted up to a hearing in July 2003. The court found that mother’s statements that she never had used methamphetamine and that she had been clean and sober for two-and-a-half years as of July 2003 were clearly false. In addition, the trial court found the following:
“Mother’s abuse of OxyContin has clearly interfered with her ability to parent [child]. Mother’s inability to be honest about the severity and the recency of her drug abuse leaves the court with grave concerns about her ability to presently parent the child.”
The trial court found that mother had been maintaining suitable housing for some time and seemed to be able to take care of the home and herself without the assistance of others. The court also found that mother had completed some parenting classes, regularly attended church, and made an effort to visit child in Portland. Based on those factors, the court found that mother had made an increased effort to change her circumstances after her hospitalization in Idaho in February 2002. Nonetheless, the court concluded that the efforts that mother had made would not significantly improve her ability to parent. That conclusion was based on the addiction counselor’s May 2002 statement that mother did not understand the disease concept, on Sweet’s July 2002 evaluation of mother, and on an August 2002 DHS report that mother did not recognize the impact of her choices and parenting skills on child, as evidenced by her statement that, while she might consider relinquishing her parental rights to her sister, she would fight forever to keep child from remaining with the Stubblefields.
The court made additional findings concerning the visitations, observations about mother’s tendency to paren-tify child and her statements to the effect that child’s concern for mother was normal and showed her compassion, and various witness’s impressions about child’s progress, including Sweet’s impression that child was much improved by the time he last saw her.
Based on the foregoing, the trial court concluded that mother is addicted to controlled substances, that she is [94]*94in denial of that fact, and that that addiction, coupled with mother’s mental illness, “has resulted in behavior that is seriously detrimental to [child],” including causing child to act as the parent in the relationship and thereby inflicting her with trauma and fear. The court stated, “Now that [child] does not have to bear that burden she is experiencing improvement in her mental health. ** * * [T]he time to reunite [child] with her mother is long gone.” Finally, the court concluded that mother’s condition is unlikely to change and that integration into mother’s home within a reasonable time is not possible and, therefore, that mother’s parental rights should be terminated.8
Mother appealed the trial court’s judgment to the Court of Appeals. That court affirmed without opinion. State ex rel Dept. of Human Services v. Simmons, 196 Or App 787, 106 P3d 699 (2004) (Simmons I). Mother petitioned this court for review and this court vacated the Court of Appeals decision and remanded the case to that court for reconsideration in light of this court’s decisions in Stillman, 333 Or 135, and State ex rel Dept. of Human Services v. Smith, 338 Or 58, 106 P3d 627 (2005). State ex rel Dept. of Human Services v. Simmons, 338 Or 374, 110 P3d 113 (2005) (Simmons II). On remand, the Court of Appeals again affirmed the judgment of the trial court terminating mother’s parental rights. Simmons III, 203 Or App 279. We discuss the specific reasons [95]*95for the Court of Appeals decision in Simmons III and this court’s holdings in Stillman and Smith below.
Under ORS 419B.500, a court may terminate parental rights if the court finds that it is in the child’s best interest to do so, but only if the state also establishes the existence of one or more statutory grounds for termination of parental rights by clear and convincing evidence. See Smith, 338 Or at 79 (stating rule); ORS 419B.521 (“[t]he facts on the basis of which the rights of the parents are terminated, unless admitted, must be established by clear and convincing evidence”). As this court stated in Smith, under that standard, the court must find that the evidence establishes that the truth of the facts asserted is highly probable. Smith, 338 Or at 79.
The specific bases for termination of parental rights are set out in ORS 419B.502 to 419B.508 and include termination for extreme parental misconduct, unfitness, neglect, and abandonment. In the present case, the state proceeded under ORS 419B.504, which provides:
“The rights of the parent or parents may be terminated as provided in ORS 419B.500 if the court finds that the parent or parents are unfit by reason of conduct or condition seriously detrimental to the child or ward and integration of the child or ward into the home of the parent or parents is improbable within a reasonable time due to conduct or conditions not likely to change. In determining such conduct and conditions, the court shall consider but is not limited to the following:
“(1) Emotional illness, mental illness or mental deficiency of the parent of such nature and duration as to render the parent incapable of providing proper care for the child or ward for extended periods of time.
“(2) Conduct toward any child of an abusive, cruel or sexual nature.
“(3) Addictive or habitual use of intoxicating liquors or controlled substances to the extent that parental ability has been substantially impaired.
“(4) Physical neglect of the child or ward.
[96]*96“(5) Lack of effort of the parent to adjust the circumstances of the parent, conduct, or conditions to make it possible for the child or ward to safely return home within a reasonable time or failure of the parent to effect a lasting adjustment after reasonable efforts by available social agencies for such extended duration of time that it appears reasonable that no lasting adjustment can be effected.
“(6) Criminal conduct that impairs the parent’s ability to provide adequate care for the child or ward.”
(Emphasis added.)
As this court observed in Stillman, the foregoing statute
“sets out a two-part test for determining whether to terminate parental rights, both parts of which must be met before the court orders termination. First, the court must address a parent’s fitness: The court must find that the parent is ‘unfit by reason of conduct or condition seriously detrimental to the child.’ That, in turn, requires a two-part inquiry: The court must find that: (1) the parent has engaged in some conduct or is characterized by some condition; and (2) the conduct or condition is ‘seriously detrimental’ to the child. Second — and only if the parent has met the foregoing criteria — the court also must find that the ‘integration of the child into the home of the parent or parents is improbable within a reasonable time due to conduct or conditions not likely to change.’ ”
333 Or at 145. In addition, this court observed in Stillman that the focus of both parts of the test for determining a parent’s unfitness
“is on the detrimental effect of the parent’s conduct or condition on the child, not just the seriousness of the parent’s conduct or condition in the abstract. Thus, the court first must identify the parent’s conduct or condition, and then measure the degree to which that conduct or condition has had a seriously detrimental effect on the child.”
Id. at 146. Finally, this court in Stillman emphasized that a parent’s fitness must be measured at the time of the parental rights termination trial. Evidence that grounds for termination may have existed at the time that the petition was filed, without evidence that those grounds continued to exist at the [97]*97time of the trial, is insufficient to support the conclusion that a parent’s parental rights should be terminated. Id. at 148-49.
In Smith, this court elaborated on the requirement in ORS 419B.504 that a parent’s conduct be “seriously detrimental” to the child. That case involved a mother of low-average intelligence whose child was removed from her care at birth because of concerns that the mother was mentally ill. The state had alleged, among other things, that, in the years between giving birth to the child and the termination proceeding, the mother had made virtually no progress in developing parenting skills and never learned to interact with the child in any of the ways that DHS personnel thought necessary or appropriate. The state’s chief complaints in that regard were that the mother failed to make her visits with the child enjoyable for the child, that the mother did not make proper eye contact with the child, that the mother did not respond appropriately to the child’s cues, that the mother refused to read any of the parenting books provided to her, and that the mother lacked “insight” into the importance of the various skills that had been taught to her in the many parenting classes and sessions that the agency had provided her.
In evaluating mother’s conduct during the visits and any detrimental effect on the child, this court agreed that mother’s behavior was not optimal. Indeed, the court agreed that the “mother did not interact with the child at a level that would ensure that the child necessarily will experience maximum emotional development.” Smith, 338 Or at 87. Nonetheless, the court concluded that
“the deficiencies perceived here are not so severe as to implicate the standard that the statute sets out for the termination of parental rights and likely are no worse than those of thousands of Oregonians who ultimately succeed, without state intervention, in raising their children safely.
“Given mother’s limitations, perfection in parenting is not attainable (if it is for anyone), but neither is it required. For more than two years, despite her recognized intellectual and social skills deficits, mother participated regularly in programs required by DHS and religiously attended [98]*98weekly visits with the child. The testimony of DHS workers observing the visits confirmed that mother played with the child with toys, although she did not offer him a variety from among the wide assortment of educational toys available, and that the child went to mother willingly, laughed when mother took his picture, and clung to her when he felt frightened. The child never was injured or in any danger during her visits. And, the evidence was undisputed that mother had provided childcare for many children over the years with no complaints. In contending that, notwithstanding the foregoing, mother is so inadequate as a parent that her condition is seriously detrimental to the child, DHS has attempted to impose a standard on mother that the statute does not contemplate.”
Id.
With that background in mind, we turn to the present case. As noted, the state alleged in its petition to terminate mother’s parental rights in this case that the following “conditions” justified termination: (1) addictive or habitual use of controlled substances to the extent that her parenting ability is substantially impaired; (2) lack of effort or failure to maintain a suitable or stable living situation; (3) failure to present a viable plan for return of the child to mother’s care; (4) an emotional or mental illness or deficiency of such nature and duration as to render mother incapable of providing care of the child for an extended period of time; (5) physical and emotional neglect of child; (6) lack of effort to adjust circumstances, conduct, or conditions to make return of the child possible; (7) failure to effect a lasting adjustment after reasonable efforts by available social services have been made; and (8) multiple medical needs that, coupled with mother’s mental health and drug abuse problems, substantially impaired mother’s ability to care for the child.
At the outset, we observe that the fact that mother had maintained a suitable and stable living situation for well over a year by the time of the trial was undisputed. In addition, the state did not present any evidence that mother neglected the child either emotionally or physically. We therefore find that the state failed to prove either of those grounds for termination by clear and convincing evidence.
[99]*99The remaining allegations boil down to the following: that mother is unfit because her history of drug abuse, her personality disorder, and her physical illness, either alone or in combination, constitute a “condition” of the kind that may justify termination of her parental rights if we find it to be seriously detrimental to child.
A parent’s history of drug abuse, if it has been addressed successfully by the time of the termination trial, is not a “conduct or condition” that necessarily renders a parent unfit at the time of that trial. Stillman, 333 Or at 148. As noted, at the time of the trial in this case, mother successfully had completed four levels of drug treatment and a 12-step program. She had been clean and sober for 20 months; no witness even hinted that mother might have been abusing her medications after February 2002. And, according to Cross, mother had been growing ever stronger in her recovery. On this record, there is not clear and convincing evidence of a present drug abuse problem that would render mother unfit at the time of the termination trial.
Notwithstanding mother’s success in treating her drug problem, the Court of Appeals found a “real and present risk of relapse” based on mother’s “persistent dishonesty and denial about her addiction.” However, there is ample evidence in the record that, by the time of trial, mother had admitted that she was an addict, that she understood that she would always be in recovery, and that she continued regular counseling with Cross and her pastor to address that issue. The fact that mother lied about the recency of her last abuse of OxyContin and continued to deny in court that she ever had used methamphetamine9 does not undercut the evidence of mother’s undisputed success in the 20 months before the trial in remaining clean and sober, and it does not, in itself, reflect on mother’s fitness or ability to care for child. See Smith, 338 Or at 85-86 (in absence of clear and convincing evidence that mother’s tendency to lie and tell outrageous [100]*100stories is seriously detrimental to child, mother’s untruthfulness not grounds for termination). The Court of Appeals erred in not focusing on the pertinent time frame. We find that the record does not contain clear and convincing evidence of a serious risk of relapse, such as to constitute a “condition” that might render mother unfit.
We also do not view mother’s physical illnesses as a condition that might justify termination. By the time of trial, and for at least a year before that, the undisputed evidence was that mother was much healthier than she ever had been before. Indeed, the trial court found that, by the time of the trial, mother had been able to take care of herself and her home without the assistance of others.
Finally, Sweet diagnosed mother with an unspecified personality disorder with borderline, dependent and narcissistic features, and Condon’s findings, although not resulting in a diagnosis, were not inconsistent with Sweet’s. Both psychologists found that mother was inflexible, over-controlled, self-centered, and overly desirous of appearing in a positive light. Sweet testified that that personality disorder would interfere with mother’s ability to interact with others and to place child’s needs above her own.10 We agree that mother’s personality disorder is a “condition” of the kind that may justify termination under ORS 419B.504, if there is clear and convincing evidence that that condition is seriously detrimental to child.
Our de novo review of the record shows the following evidence concerning the effect of mother’s “condition” on child: Immediately after child was removed from mother’s care, child was anxious and upset and suffered post-traumatic stress over having witnessed mother’s drug abuse [101]*101and having been unable on occasion to wake mother up, leading her to worry that mother would die, and over having witnessed repeated, loud arguments between mother and Dahlem. At that time, child also was irritable, angry, and defiant. Witnesses agreed that child was “parentified,” insofar as she was inordinately worried about her mother’s welfare and felt herself to be responsible for her mother. She began to wet the bed at night and wet herself occasionally during the day. In her play, child exhibited a fear of abandonment.
By the time of the trial, however, child was relatively happy and well adjusted. She had friends and did well in school, and her enuresis was resolved. Portland, who had met with child three times in the weeks before the termination trial, testified that she had seen marked improvement in child’s behavior and outlook during the time that child was in foster care. According to Portland, child “is happy. There’s sparkle in her eyes. She has friends she cares about. * * * Emotionally, she can talk about her feelings, negative and positive.” And, with respect to child’s “parentification,” child’s therapists had addressed child’s feelings of responsibility for mother together with mother, mother has accepted responsibly for the problems, and child now knows that mother loves her and that she is not responsible for mother.
Portland also testified, however, that child continues to display an inability to trust others and a tendency to tell others what she thinks they want to hear. Finally, the record reflects that, immediately before the trial in this matter, child was beginning to be inattentive and was showing off and disorganized at school, although Portland acknowledged that those behaviors likely were brought on by anxiety related to the impending proceeding.
With respect to mother’s interactions with child, virtually all witnesses agreed that mother and child love each other very much and that mother and child were extremely loving and affectionate with each other during visitations.11 Mother played with child, brought age-appropriate activities [102]*102to the visits, brought and prepared nutritious meals for child, read child books, and disciplined child appropriately as necessary.
The state’s chief criticisms of mother were that she did not fully appreciate the harm that child had suffered from witnessing mother’s drug abuse and arguments with Dahlem and that mother was not able to put child’s needs above her own. As to the former point, there is evidence in the record that, soon after child was removed from mother’s care, mother minimized child’s fears and attempted to force contact with Dahlem. The record does not support a finding, however, that that attitude persisted after mother stopped abusing drugs and began her recovery in earnest in February 2002. The evidence in support of that latter point included the following: (1) mother believed that child’s worries for mother’s well-being showed child’s compassion and were not problematic; (2) mother disregarded child’s wishes on a number of occasions, including not permitting child to cut her hair, giving a puppy a name different from the one child preferred, and writing a message to child in sidewalk chalk in spite of child’s request that she not do so; and (3) mother brought presents and played with child during visits rather than talking about the family issues.
The foregoing evidence of detriment to the child does not rise to the level contemplated by the legislature in ORS 419B.504 so as to provide a basis for concluding that mother is unfit. Child’s problems, such as they continue to exist, do not appear to us to be severe; all agree that child presently is generally happy and well adjusted.12 And mother’s behavior [103]*103before the termination trial, during the period of her recovery, while not exemplary in every respect, was mainly positive and appropriate. During visits, mother consistently was loving and appropriate toward child and child freely hugged and kissed mother and told mother that she loved her. Additionally, there is no evidence in the record that child ever was harmed or in danger during that period. As this court stated in Smith, 338 Or at 87, perfection in parenting is neither attainable nor required.
None of the evidence that the state presented pertaining to mother’s parenting after child was removed, and particularly after mother began her recovery in earnest in February 2002, shows clearly and convincingly that mother’s parenting is so inadequate as to be seriously detrimental to child, thereby justifying termination of mother’s parental rights. Certainly, there is no clear and convincing evidence that such was the case at the time of trial.13 Again, as in Smith, the state is attempting to impose a standard of parenting on mother that the statute does not contemplate. Smith, 338 Or at 87. Based on the foregoing, we conclude that the state has failed to show by clear and convincing evidence that mother is unfit under ORS 419B.504. It follows that the trial court order terminating mother’s parental rights was error, as was the Court of Appeals decision affirming that order.
The decision of the Court of Appeals is reversed. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings.