JUSTICE GABRIEL delivered the Opinion of the Court.
¶1 This case requires us to decide whether the trial court abused its discretion in permitting a police officer to testify regarding the results of a Horizontal Gaze Nystagmus ("HGN") test without first qualifying that officer as an expert witness under CRE 702 and Venalonzo v. People , 2017 CO 9, 388 P.3d 868.1
¶2 We conclude that, on the facts of this case, the officer's testimony concerning the HGN test was expert testimony under CRE 702 and that the district court therefore erred in holding otherwise. We further conclude, however, that on the facts presented here, the court's error in admitting the testimony was harmless.
¶3 Accordingly, we affirm the district court's judgment.
I. Facts and Procedural History
¶4 One evening at around midnight, a police officer was on patrol when he noticed a truck drive past him with only one headlight on. Because of that, the officer pulled the truck over and initiated a discussion with the driver, Randy Campbell. The only other person in the car was a friend of Campbell's, who was intoxicated and sitting in the front passenger seat.
¶5 In speaking with Campbell, the officer detected a moderate odor of alcohol on Campbell's breath and noted that his eyes were bloodshot and his speech was slightly slurred. In addition, when the officer asked to see Campbell's identification and Campbell tried to remove it from his wallet, he dropped the wallet on the floor, although he was eventually able to retrieve the identification for the officer.
¶6 Based on these observations, the officer asked Campbell if he had been drinking. Campbell initially responded that he had not had any alcohol that evening, but after the *74officer said that he could smell alcohol on Campbell, Campbell responded that he had had one beer at approximately 2:30 p.m. that afternoon. The officer then requested that Campbell perform a series of voluntary roadside sobriety maneuvers, and Campbell agreed to do so. According to the officer, when Campbell attempted to get out of the truck, he unsuccessfully reached for the door handle twice before finally grabbing it.
¶7 The officer proceeded to administer three roadside sobriety maneuvers, namely, the HGN test, the "walk and turn," and the "one-leg stand," all of which the officer had been trained on and was certified to administer. After determining that Campbell had failed to perform all three of the maneuvers as a sober person would, and based on the officer's observations of Campbell and on Campbell's admission that he had been drinking, the officer arrested him.
¶8 Following the arrest, the officer noticed three small bottles of alcohol-one opened and two unopened-on the floor of Campbell's truck. In addition, with Campbell's consent, the officer administered two breath tests to determine Campbell's blood alcohol content ("BAC"). These tests showed that Campbell had a BAC of 0.07 and 0.086, respectively, which allowed a permissible inference that Campbell was, at a minimum, impaired by the consumption of alcohol. See § 42-4-1301(6)(a)(II)-(III), C.R.S. (2018).
¶9 The prosecution subsequently charged Campbell with driving under the influence ("DUI"), driving while ability impaired ("DWAI"), operating a vehicle with a defective headlamp, and possessing an open alcohol container. Campbell pleaded not guilty and demanded a jury trial.
¶10 The case proceeded, and during pretrial proceedings, Campbell noted that the prosecution intended to have the officer who conducted the roadside maneuvers on him testify as a lay witness. Campbell argued that such maneuvers comprised "a specialized area of knowledge" and therefore the officer should only be permitted to testify about his observations and should not be allowed to offer "any opinions that stem from those observations that are based on this specialized knowledge that the officer had."
¶11 The court responded that it was working in a bit of a vacuum because it did not know what the officer's testimony would be. The court observed that there could be testimony on roadsides that would require expert opinion but that general testimony by an officer that he was trained to do the test in a certain way, that he did so, and that people who are intoxicated look a certain way while people who are sober look a different way was not scientific testimony necessitating any expert designation. The court then stated that it was not making a finding one way or another but that it would consider an objection at the appropriate time. The case then proceeded to trial.
¶12 At trial, the prosecution called the officer as a witness, and the officer testified at length about his training and experience regarding the roadside maneuvers that he had conducted on Campbell and regarding the maneuvers themselves. In particular, with respect to his qualifications, the officer testified that (1) he was "certified in the standard field sobriety roadside maneuvers and DUI detection"; (2) to obtain this certification, he had taken an initial course at the police academy, which he stated may have been forty hours in length (although he was unsure of that); and (3) he was subsequently recertified every other year through a sixteen-hour course. In addition, he estimated that he had performed approximately seven hundred DUI investigations during his then-fifteen-years of service as a police officer, and he explained that in his training, he had gone through what officers called the "wet lab." In the wet lab, trainees administered the standardized maneuvers on test subjects to see if the subjects could perform the maneuvers as a sober person would. The officer confirmed that at the time he conducted the voluntary roadside maneuvers on Campbell, he was certified to administer them.
¶13 The officer then testified that he typically conducted three different sobriety maneuvers on someone whom the officer suspected was driving under the influence and that he did so in this case. With respect to the HGN test in particular, the officer explained that "you're looking for a total of six *75clues, three separate clues in each eye." He then described these clues in detail:
During the initial part of it ..., what you do is you hold a stimulus, which is my finger approximately 12 inches from the individual's nose. You then instruct them to follow that stimulus with their eyes, and their eyes only[,] keeping their head perfectly still. As you move it or start the maneuver you start with the left side of the body and what you're initially looking for is equal tracking in the individual's eyes. ... The other thing that you're looking for is the pupil size. ... The next clue or when you start into the clues is lack of smooth pursuit. And what you're looking for is[,] as the eyes are following the stimulus[,] whether or not the eye is stopping or if it's following the stimulus smoothly as it passes in front of their face. The second clue that you're looking for is distinct and sustained nystagmus at maximum deviation and what you do there is you hold the stimulus out at maximum deviation[,] which is basically there's no white left in the corner of the eye[,] and what you're looking for is an involuntary jerkiness of the eye[;] the best way to describe it is if you take a basketball and bounce it against the side of the wall[,] that's the kind of the movement that the eye would make. And the third clue that you're looking for is an onset of nystagmus prior to 45 degrees.
¶14 The officer stated that in this case, he "observed all six clues on the Defendant" and specifically that he "observed the lack of smooth pursuit," "a distinct and sustained nystagmus in both eyes," and "a[n] onset of nystagmus prior to 45 degrees in both eyes."
¶15 Based on the foregoing testimony, the prosecution asked the officer, "Now [was Campbell's] performance of the HGN[,] in your opinion[,] based on your training and experience[,] was that consistent with a sober person?" The officer responded, "No, it was not." At this point, defense counsel objected that the officer's statements amounted to expert testimony, but the court overruled that objection.
¶16 Notably, at trial, Campbell did not deny that he had been drinking. His theory of defense was that he drank after he was pulled over. His counsel thus told the jury that "the key detail here is when Mr. Campbell drank. He drove, then he drank." She went on to explain that Campbell "took a shooter of schnapps [after being pulled over] to calm his nerves for the interaction with the officer." Counsel suggested that, consequently, even though Campbell was impaired when he took the sobriety and breath tests, he was not impaired while driving, as evidenced by the alleged fact that he had not been driving poorly or weaving all over the road. In further support of Campbell's defense, counsel introduced testimony from some of Campbell's friends and from his son, all of whom testified that they had been in Campbell's company in the hours before Campbell set out in his car and that they had not seen him consume any alcohol.
¶17 The jury ultimately rejected Campbell's assertions and convicted him of DWAI, possessing an open alcohol container, and operating a vehicle with a defective headlamp.
¶18 Campbell appealed to the Arapahoe County District Court, but that court affirmed. As pertinent here, the court determined that with respect to the HGN test, the officer developed his opinions by watching Campbell's eye movement, which the court opined was something that an ordinary citizen could do. The court thus concluded that the officer's testimony "was admissible under C.R.E. 701 as lay witness testimony." The court further stated that any trial errors were harmless, although the court did not explain or elaborate on that statement.
¶19 Campbell then sought, and we granted, certiorari review.
II. Analysis
¶20 We begin by setting forth the standard of review governing a trial court's evidentiary decisions. We then discuss our prevailing law regarding expert and lay testimony, and we apply those principles to the facts of this case, concluding that the officer's testimony here amounted to expert testimony that was improperly admitted as lay testimony. Having made this determination, we proceed to *76consider whether this error was harmless, and we conclude that it was.
A. Standard of Review
¶21 We review a trial court's evidentiary rulings for an abuse of discretion. Venalonzo , ¶ 15, 388 P.3d at 873. A trial court abuses its discretion when its ruling is manifestly arbitrary, unreasonable, or unfair. Id .
¶22 If we determine that a trial court has abused its discretion regarding a preserved, nonconstitutional issue, then we must consider whether the error was harmless. Romero v. People , 2017 CO 37, ¶ 16, 393 P.3d 973, 978. Under this standard, reversal is required only if the error affected the parties' substantial rights. Hagos v. People , 2012 CO 63, ¶ 12, 288 P.3d 116, 119 ; see also C.A.R. 35(c) ("The appellate court may disregard any error or defect not affecting the substantial rights of the parties."); Crim. P. 52(a) ("Any error, defect, irregularity, or variance which does not affect substantial rights shall be disregarded."); CRE 103(a) ("Error may not be predicated upon a ruling which admits or excludes evidence unless a substantial right of the party is affected[.]"). Thus, we will reverse only if the error substantially influenced the verdict or impaired the fairness of the trial. Hagos , ¶ 12, 288 P.3d at 119.
B. The Officer's Testimony Was Expert Testimony
¶23 Campbell principally contends that the trial court abused its discretion in admitting the officer's testimony regarding the HGN test as lay testimony. We agree.
¶24 Under CRE 701, a lay witness's testimony in the form of opinions or inferences is limited to those opinions or inferences that are "(a) rationally based on the perception of the witness, (b) helpful to a clear understanding of the witness' testimony or the determination of a fact in issue, and (c) not based on scientific, technical, or other specialized knowledge within the scope of Rule 702." CRE 702, in turn, provides, "If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise."
¶25 In Venalonzo , ¶ 2, 388 P.3d at 870-71, we articulated the standard to be applied in distinguishing lay from expert testimony. As pertinent here, we held that
in determining whether testimony is lay testimony under Colorado Rule of Evidence ("CRE") 701 or expert testimony under CRE 702, the trial court must look to the basis for the opinion. If the witness provides testimony that could be expected to be based on an ordinary person's experiences or knowledge, then the witness is offering lay testimony. If, on the other hand, the witness provides testimony that could not be offered without specialized experiences, knowledge, or training, then the witness is offering expert testimony.
Id.
¶26 Applying these principles to the present facts, we conclude, for several reasons, that the officer's testimony was expert testimony.
¶27 First, the officer's testimony regarding the HGN test was not the type of testimony that could be offered without specialized experiences, knowledge, or training. As noted above, the officer testified in extensive detail, using specialized and technical terminology, about how the HGN test is administered, the clues that he looks for, and how he interprets those clues. Although the People contend otherwise, we perceive no basis for concluding that such testimony could be expected to be based on an ordinary person's experiences or knowledge.
¶28 Indeed, as described above, the officer himself testified regarding the extensive training that he undertook and the accreditations that he had received, which prepared and qualified him to administer and interpret the HGN test. He further testified regarding his extensive experience in the field, noting that, as of the time of trial, he (1) had fifteen years of experience as a police officer; (2) had conducted approximately seven hundred DUI investigations; and (3) was certified to administer *77each of the voluntary roadside maneuvers that he had performed here, including the HGN test.
¶29 Second, the prosecution relied on the officer's extensive and specialized experiences, knowledge, and training, specifically asking the officer, "Now [Campbell's] performance of the HGN in your opinion[,] based on your training and experience [,] was that consistent with a sober person?" (Emphasis added.) This question underscored that the officer's testimony was not the type of testimony that someone with only a lay understanding could have been expected to offer.
¶30 Third, although the People argue that anyone who has undergone an eye examination would understand the HGN test, we perceive nothing in the record to support such a claim. Nor can we agree that a lay person without training would be familiar with the principles of the HGN test and would have been able to explain the administration and interpretation of that test to the jury as the officer did.
¶31 For these reasons, we conclude that, on the facts of this case, the officer provided testimony regarding the HGN test that could not have been offered "without specialized experiences, knowledge, or training" and that therefore, the officer's testimony amounted to expert testimony under CRE 702. See Venalonzo , ¶ 2, 388 P.3d at 870-71 ; see also Romero , ¶ 15, 393 P.3d at 977-78 (concluding that the trial court abused its discretion when it allowed a police officer to testify as a lay witness on the concept of "grooming" in a case of sex assault on a child because an ordinary citizen could not have been expected to have the experience, skills, or knowledge required to understand that concept as it relates to sexual predation); People v. Ramos , 2017 CO 6, ¶¶ 9-10, 388 P.3d 888, 891 (concluding that a detective's testimony regarding the difference between blood cast-off and blood transfer evidence was expert testimony because (1) an ordinary citizen would not be expected to have the experience, skills, or knowledge to differentiate between the two; (2) in making the distinction between the two, the officer had relied on his extensive experience as both a police officer and a detective, having worked on thousands of cases involving blood; (3) the People capitalized on the detective's specialized experience and knowledge by asking questions that invoked his training and experience; and (4) the detective used technical terms in offering his opinion).
¶32 Accordingly, we conclude that the trial court abused its discretion when it admitted the officer's testimony as lay testimony and did not require the officer to be qualified as an expert witness.
¶33 This determination, however, does not end our analysis because we must still consider whether the trial court's error in admitting this expert testimony was harmless. We turn next to that question.
C. The Trial Court's Error Was Harmless
¶34 In this case, the parties appear to agree that if the trial court erred in allowing the officer to testify as a lay witness regarding the HGN, the issue was preserved and was nonconstitutional. Accordingly, we review the error under the nonconstitutional harmless error standard, and we will reverse only if the error affected Campbell's substantial rights. Hagos , ¶ 12, 288 P.3d at 119.
¶35 Here, for several reasons, we conclude that the trial court's error in admitting the officer's expert testimony was harmless.
¶36 First, the evidence of Campbell's intoxication was overwhelming. As noted above, when the officer spoke with Campbell, he noted that Campbell's breath had an odor of alcohol, his eyes were bloodshot, his speech was slurred, and he dropped his wallet while trying to retrieve his identification. In addition, when Campbell tried to get out of his truck, the officer saw him reach for the door handle twice without success before grabbing it and opening the door.
¶37 Second, even setting aside the evidence concerning the HGN test, Campbell failed each of the other two roadside sobriety maneuvers that the officer had administered, namely, the "walk and turn," which the officer explained examines whether a person can perform multiple tasks at once, and the "one-leg stand," which tests balance.
*78¶38 Third, following Campbell's arrest, the officer found three small bottles of alcohol, one opened and two unopened, on the floor of Campbell's car.
¶39 Fourth, with Campbell's consent, the officer administered two breath tests, and those tests revealed a BAC of 0.07 and 0.086, respectively, which, as noted above, established a permissible inference that Campbell was, at a minimum, impaired by the consumption of alcohol.
¶40 Finally, Campbell admitted that he had consumed alcohol before the officer questioned him, and he did not deny that he was impaired when the officer administered the sobriety and breath tests. He asserted, however, that he drank and became impaired in the short time between when the officer pulled him over and when the officer first spoke with him. Although Campbell attempted to support this theory by calling witnesses to testify that they had not seen him drink earlier, several of the witnesses had themselves been intoxicated that day, and others could not have known whether or how much alcohol Campbell may have consumed prior to driving.
¶41 In our view, the foregoing evidence overwhelmingly supported the jury's determination that Campbell drove while his ability was impaired by alcohol, and therefore, we conclude that any error here was harmless. See People v. Stewart , 55 P.3d 107, 124-25 (Colo. 2002) (concluding that the trial court's error in admitting a police officer's testimony as lay opinion was harmless because the officer's testimony was either corroborated by the defendant or by other witnesses or was undisputed and therefore could not be said to have substantially influenced the verdict or to have impaired the fairness of the trial); Tevlin v. People , 715 P.2d 338, 342 (Colo. 1986) (concluding that the trial court's erroneous admission of certain expert testimony was harmless "[i]n light of the overwhelming evidence of guilt produced in th[e] case").
¶42 In concluding that the trial court's error here was harmless, we are unpersuaded by Campbell's argument that the HGN testimony was not harmless because it was "key to the prosecution's case" and influenced the jury's verdict, and because the officer's testimony regarding the HGN test was "the most difficult to challenge." In our view, these assertions, even if true, are insufficient to overcome the overwhelming evidence described above.
¶43 Accordingly, we conclude that the trial court's error in allowing the officer to testify about the HGN test without qualifying him as an expert witness was harmless and does not warrant reversal.
III. Conclusion
¶44 For the foregoing reasons, we conclude that the officer's testimony regarding his administration of and conclusions regarding the HGN test was testimony that could not have been offered without specialized experiences, knowledge, or training and that, therefore, this testimony was expert testimony. Accordingly, we further conclude that the trial court abused its discretion in admitting this testimony without first qualifying the officer as an expert witness. In light of the overwhelming evidence of Campbell's guilt, however, we ultimately conclude that any error by the trial court in admitting the officer's testimony as lay testimony was harmless.
¶45 Accordingly, we affirm the judgment below.