Opinion
WERDEGAR, J.
In this action for medical monitoring of the residents of a geographic area affected by defendants’ toxic chemical discharge, the question before us is whether plaintiffs, in moving for class certification, have met their burden of demonstrating that common issues of law and fact predominate. We conclude they have not. We therefore affirm the judgment of the Court of Appeal.
Background
Plaintiffs Roslyn Carrillo et alia allege that defendants Lockheed Martin Corporation et alia, in the course of conducting manufacturing operations in the City of Redlands, beginning in 1954, discharged dangerous chemicals that contaminated the city’s drinking water with harmful toxins and that this contaminated water was used by a large portion of the city’s residents. In December 1996, on behalf of themselves and persons similarly situated, plaintiffs filed this action in the San Bernardino County Superior Court.
[1102]*1102Plaintiffs pray that the court order defendants to fund a court-supervised program for the medical monitoring of class members, and for punitive damages.
Plaintiffs moved for certification of a “medical monitoring” class and a “punitive damage” class, defined identically as “People who were exposed to water contaminated with any of the following chemicals: TCE, PCE, TCA, other solvents, Ammonium Perchlorate, Perchlorate, other unknown rocket fuel components and rocket fuel decomposition products, Beryllium, Carbon Tetrachloride, Vinyl Chloride, Hydrazine (and Hydrazine derivatives), Nitrosamines (and Nitrosamine derivatives), Epoxides (and Epoxide derivatives), Triazines (and Triazine derivatives), at levels at or in excess of the dose equivalent of the MCE (Maximum Contaminant Level),[1] or in excess of the safe dose where there is no MCL, for some part of a day, for greater than 50% of a year, for one or more years from 1955 to the present” within specified geographical limits. (Fns. omitted.) Plaintiffs’ class definition indicated that review of relevant water quality documents was ongoing and that the definition would be amended if additional chemicals were identified.
One of plaintiffs’ attorneys declared that estimating the number of persons in the class was difficult, because the University of Redlands is located within the specified geographic boundaries, and persons residing, working or studying within the defined area may qualify as class members. The attorney’s best estimate was that the class includes between 50,000 and 100,000 people.
The trial court certified the classes, finding that plaintiffs had met their burden of proof under Code of Civil Procedure section 382: “The Court finds that the plaintiffs have a realistic chance of success on the merits, [^f] Specifically, the Court finds that the plaintiffs have shown that there is a realistic chance that the defendants caused contaminants to be leaked into the water table beneath Redlands and that this contaminated water was served to the members of the proposed class.” The court also found that there is an ascertainable class, concluding it was “not necessary to determine the levels of toxins received by each plaintiff at this time and that the geographic limits placed on the class are reasonable and related to the alleged contamination.” The court concluded, finally, that members of the class have a well-defined community of interest and that common questions of law and fact predominate in the action.
Parties objecting to certification filed three writ petitions in the Court of Appeal, which that court consolidated. Opining that individual issues raised [1103]*1103by plaintiffs’ claims “clearly predominate, making class certification inappropriate,” the Court of Appeal granted a writ of mandate directing the trial court to vacate its order certifying the classes. We granted plaintiffs’ petition for review.
Discussion
I. Suitability of Medical Monitoring Claims for Class Treatment2
We first addressed the availability of medical monitoring as a form of damages in Potter v. Firestone Tire & Rubber Co. (1993) 6 Cal.4th 965 [25 Cal.Rptr.2d 550, 863 P.2d 795] (Potter). There, residents of homes located near a landfill at which the dumping of toxic substances was prohibited brought, as individual claimants, an action against a tire manufacturing company that had dumped toxic waste materials, alleging that their water supply had thereby been contaminated. The plaintiffs sought damages for, inter alia, fear of cancer and the costs of medical monitoring. (See id. at pp. 975-979.) Recognizing that “ ‘expenditures for prospective medical testing and evaluation, which would be unnecessary if the particular plaintiff had not been wrongfully exposed,’ ” are “ ‘detriment proximately caused’ ” by negligent disposal of toxic substances (id. at p. 1005 & fn. 24, quoting Civ. Code, § 3333), we held that “the cost of medical monitoring is a compensable item of damages where the proofs demonstrate, through reliable medical expert testimony, that the need for future monitoring is a reasonably certain consequence of a plaintiff’s toxic exposure and that the recommended monitoring is reasonable” (Potter, supra, at p. 1009).
“In determining the reasonableness and necessity of monitoring,” we stated, “the following factors [(hereafter the Potter factors)] are relevant: (1) the significance and extent of the plaintiffs exposure to chemicals; (2) the toxicity of the chemicals; (3) the relative increase in the chance of onset of disease in the exposed plaintiff as a result of the exposure, when compared to (a) the plaintiffs chances of developing the disease had he or she not been exposed, and (b) the chances of the members of the public at large of developing the disease; (4) the seriousness of the disease for which the plaintiff is at risk; and (5) the clinical value of early detection and diagnosis.” (Potter, supra, 6 Cal.4th at p. 1009.)
We have not previously addressed the prerequisites for class treatment of medical monitoring claims. “Section 382 of the Code of Civil Procedure authorizes class suits in California when ‘the question is one of a common or general interest, of many persons, or when the parties are [1104]*1104numerous, and it is impracticable to bring them all before the court.’ The burden is on the party seeking certification to establish the existence of both an ascertainable class and a well-defined community of interest among the class members.” (Washington Mutual Bank v. Superior Court (2001) 24 Cal.4th 906, 913 [103 Cal.Rptr.2d 320, 15 P.3d 1071] (Washington Mutual).)3
Plaintiffs assert that separate litigation of each class member’s medical monitoring claim would unnecessarily consume vast judicial resources and time. They also urge us to repudiate the Court of Appeal’s suggestion that the presence of individual issues generally precludes class certification in mass toxic exposure cases, arguing any such categorical foreclosure would render our decision in Potter meaningless. Defendants, on the other hand, emphasize that Potter's proximate cause rationale for recognizing medical monitoring costs as damages logically extends only to such “increased or different monitoring” (Potter, supra, 6 Cal.4th at p. 1009, fn. 27) as a defendant’s conduct actually necessitates. In light of their due process right to litigate each individual plaintiffs actual toxic dosage and relevant personal characteristics, defendants argue, individual issues in the case predominate over common ones, such that the community of interest required for class certification is lacking.
The certification question is “essentially a procedural one that does not ask whether an action is legally or factually meritorious.” (Linder v. Thrifty Oil Co. (2000) 23 Cal.4th 429, 439-440 [97 Cal.Rptr.2d 179, 2 P.3d 27] (Linder).) “The community of interest requirement [for class certification] embodies three factors: (1) predominant common questions of law or fact; (2) class representatives with claims or defenses typical of the class; and (3) class representatives who can adequately represent the class.” (Richmond v. Dart Industries, Inc. (1981) 29 Cal.3d 462, 470 [174 Cal.Rptr. 515, 629 P.2d 23].) Plaintiffs acknowledge it is their burden to establish the requisite community of interest and that “the proponent of certification must show, inter alia, that questions of law or fact common to the class predominate over the questions affecting the individual members.” (Washington Mutual, supra, 24 Cal.4th at p. 913.)
“The ultimate question in every case of this type is whether ... the issues which may be jointly tried, when compared with those requiring [1105]*1105separate adjudication, are so numerous or substantial that the maintenance of a class action would be advantageous to the judicial process and to the litigants.” (Collins v. Rocha (1972) 7 Cal.3d 232, 238 [102 Cal.Rptr. 1, 497 P.2d 225]; see also Linder, supra, 23 Cal.4th at p. 435.)
Defendants point to our statement that the Potter factors comprise “substantial evidentiary burdens” for plaintiffs (Potter, supra, 6 Cal.4th at p. 1009), as if to suggest the factors constitute novel proof requirements applicable only in medical monitoring cases. Not so. Potter recognizes “not a separate tort but simply an item of damages that cannot be awarded until liability is established under a traditional tort theory.” (San Diego Gas & Electric Co. v. Superior Court (1996) 13 Cal.4th 893, 914, fn. 18 [55 Cal.Rptr.2d 724, 920 P.2d 669], citing Potter, supra, at pp. 1006-1007.) So to observe does not gainsay the high court’s observation that “limitations and cautions [like the Potter factors are] important—and integral—parts of the state-court decisions that permit asymptomatic plaintiffs a separate tort claim for medical monitoring costs.” (Metro-North Commuter R. Co. v. Buckley (1997) 521 U.S. 424, 444 [117 S.Ct. 2113, 2124, 138 L.Ed.2d 560]; see also id. at pp. 440-441 [117 S.Ct. at pp. 2122-2123], citing, inter alia, Potter, supra, at p. 1010, fn. 28.)
As defendants acknowledge, Potter simply specified for the medical monitoring context the traditional requirement that a plaintiff prove causation of damage. Thus, while in Potter we “ma[de] it clear that the monitoring must be ‘additional or different’ ” than that previously required (Gutierrez v. Cassiar Mining Corp. (1998) 64 Cal.App.4th 148, 156 [75 Cal.Rptr.2d 132]), we just as clearly stated that “if additional or different tests and examinations are necessitated as a result of the toxic exposure caused by the defendant, then the defendant bears full responsibility for their costs” (Potter, supra, 6 Cal.4th at p. 1012, fn. 31).
Defendants assert that “the required proof under Potter” includes “that each of the elements of the claims asserted on behalf of proposed class members, and all applicable defenses, are capable of common proof.” Again, not so. We consistently have recognized, before and after Potter, that “the fact that each member of the class must prove his [or her] separate claim to a portion of any recovery by the class is only one factor to be considered in determining whether a class action is proper.” (Vasquez v. Superior Court (1971) 4 Cal.3d 800, 809 [94 Cal.Rptr. 796, 484 P.2d 964, 53 A.L.R.3d 513].)4
In sum, no per se or categorical bar exists to a court’s finding medical monitoring claims appropriate for class treatment, so long as any individual [1106]*1106issues the claims present are manageable. Accordingly, we shall review the certification ruling before us in light of the established standards for class certification generally.
II. Plaintiffs Demonstrated Presence of Some Common Issues 5
As indicated, in granting plaintiffs’ certification motion, the trial court expressly found that common questions predominate and determined that any individual issues that might arise at the time of trial are manageable. “Because trial courts are ideally situated to evaluate the efficiencies and practicalities of permitting group action, they are afforded great discretion in granting or denying certification.” (Linder, supra, 23 Cal.4th at p. 435.) Nevertheless, “we must examine the trial court’s reasons for [granting] class certification.” (Id. at p. 436; see also Washington Mutual, supra, 24 Cal.4th at p. 914.) In particular, we must consider whether the record contains substantial evidence to support the trial court’s predominance finding, as a certification ruling not supported by substantial evidence cannot stand. (Linder, supra, at pp. 435-436; see also Richmond v. Dart Industries, Inc., supra, 29 Cal.3d at p. 470.)
At the outset, the record reveals that plaintiffs’ claims sound generally in negligence, entailing proof of the “well-known elements of any negligence cause of action, viz., duty, breach of duty, proximate cause and damages.” (Artiglio v. Corning Inc. (1998) 18 Cal.4th 604, 614 [76 Cal.Rptr.2d 479, 957 P.2d 1313].)6 Addressing whether questions common to the class predominate over questions affecting members individually, therefore, required the trial court to consider these elements.
Whether defendants in disposing of their chemical wastes owed a duty of care to the class members, i.e., to the persons who lived for the specified period within the specified geographical area, is a question of law for the court. (Parsons v. Crown Disposal Co. (1997) 15 Cal.4th 456, 472 [63 Cal.Rptr.2d 291, 936 P.2d 70].) Defendants proffer no reason why a court would need to engage in individualized analysis in order to answer that question. The trial court rationally could conclude that the duty element of plaintiffs’ claims will be susceptible of common proof.
[1107]*1107Additionally, how and when defendants disposed of toxic chemicals and whether defendants’ conduct was negligent are, as the Court of Appeal recognized, significant common issues of fact in this case. The parties already have presented extensive evidence (including well sampling and other hydrological data) about the pattern and degree of contamination of Redlands groundwater with various chemicals and the potential health consequences to humans of exposure to those chemicals. Defendants have conceded that common issues are present in the case because defendants’ acts allegedly are the same with regard to each plaintiff. Thus, the record also contains substantial evidence supporting the conclusion that the breach of duty element of plaintiffs’ claims will be susceptible of common proof.7
As noted, when first recognizing the medical monitoring remedy in Potter, we focused on the causation and damages elements of such claims, stating that in order to recover plaintiffs must demonstrate, through reliable medical expert testimony, both that the need for future monitoring is a “reasonably certain consequence” of toxic exposure and that the monitoring sought is “reasonable.” (Potter, supra, 6 Cal.4th at p. 1009.) Defendants take the position that plaintiffs in moving for class certification have failed to demonstrate either that the causation (“reasonably certain consequence”) or the damages (“reasonable” monitoring) elements of their medical monitoring claims will be susceptible of common proof.
Plaintiffs clearly are in a position to address some aspects of causation and damages on a class basis. Defendants concede, for example, that “the toxicity of the chemicals” allegedly discharged and “the seriousness of [any] disease for which the plaintiff is at risk”—both factors discussed in Potter, supra, 6 Cal.4th at page 1009—would be susceptible of common proof. And as the Court of Appeal noted, “the amount of contaminants that entered the groundwater; and, when, where, and at what levels were contaminants pumped by the city’s wells entered into the domestic water system” are significant common issues of fact in this case.
Plaintiffs contend that, on the theory of liability they intend to present, each individual’s exact dosage of each discharged chemical will not be relevant. According to expert testimony already in the record, plaintiffs argue, “anyone living or working in the area of contamination for at least six months has a plausible claim for medical monitoring.” Class membership, plaintiffs stress, is restricted by definition to persons who have received a specified “medically significant” minimum dosage “for some part of a day, [1108]*1108for greater than 50% of a year, for one or more years from 1955 to the present” within specified geographical boundaries. All who meet that definition, plaintiffs propose to prove, “will require a generalized monitoring program for the diseases caused by such exposure.” On such a theory, plaintiffs argue, specific individual dosages above the specified minimum are not relevant and, therefore, “the significance and extent” of toxic exposure (Potter, supra, 6 Cal.4th at p. 1009) will involve largely common proof.
The trial court in ruling on the certification motion apparently took plaintiffs’ minimum dosage liability theory into account, stating that “proof of the [actual] dosage received [by each plaintiff] is not necessary at this time.” Strictly speaking, that is correct, as in ruling on certification a court does not “ask whether [plaintiffs’] action is legally or factually meritorious.” (Linder, supra, 23 Cal.4th at pp. 439-440.) Moreover, nothing in Potter precludes liability premised on a theory that a defendant’s negligence has necessitated increased or different monitoring for all, or nearly all, exposed individuals, as long as the need is “a reasonably certain consequence of the exposure.” (Potter, supra, 6 Cal.4th at p. 1006.) That a class of water consumers could, under particularly egregious circumstances, demonstrate that everyone who drank from a polluted municipal water system over a specified period is at significant risk for having received a dose sufficient to cause serious disease and, therefore, needs special monitoring, is not inconceivable. Thus, on an appropriate theory, even dosage issues may be susceptible of common proof.
III. Plaintiffs Failed to Demonstrate Common Issues Predominate
Plaintiffs’ burden on moving for class certification, however, is not merely to show that some common issues exist, but, rather, to place substantial evidence in the record that common issues predominate. (Washington Mutual, supra, 24 Cal.4th at p. 913.) As we previously have explained, “this means ‘each member must not be required to individually litigate numerous and substantial questions to determine his [or her] right to recover following the class judgment; and the issues which may be jointly tried, when compared with those requiring separate adjudication, must be sufficiently numerous and substantial to make the class action advantageous to the judicial process and to the litigants.’ ” (Id. at pp. 913-914, quoting City of San Jose v. Superior Court (1974) 12 Cal.3d 447, 460 [115 Cal.Rptr. 797, 525 P.2d 701, 76 A.L.R.3d 1223].)
While the record on certification undoubtedly contains substantial evidence that many Redlands residents were exposed to toxic chemicals during the class period, evidence of exposure alone cannot support a finding that [1109]*1109medical monitoring is a reasonably necessary response. (Potter, supra, 6 Cal.4th at p. 1009.) As defendants emphasize, that all plaintiffs exposed to Redlands water received identical dosages of any toxic chemicals it contained is unlikely. On the one hand, duration of exposure to polluted water will vary among class members, as the class would include numerous people who lived in Redlands for a relatively short period of time during the more than 40-year class period. On the other hand, as the Court of Appeal observed, severity of exposure among class members may vary according to the amount of water they used.
Examination of the instant record reveals that plaintiffs have not provided substantial evidence that they are in a position to resolve possible dosage issues with common proof. Each class member’s actual toxic dosage would remain relevant to some degree even if plaintiffs’ “minimum dosage” liability theory ultimately were to prove viable. Membership in the class as plaintiffs have defined it requires, not merely exposure to water contaminated with one or more of the chemicals listed in the definition, but exposure “at levels at or in excess of the dose equivalent of the MCE (Maximum Contaminant Level), or in excess of the safe dose where there is no MCL” for at least the defined minimum period of time. (Fns. omitted.) But plaintiffs’ experts did not unqualifiedly opine that all who resided in Redlands for the defined period likely received such dosages. Dr. Dahlgren was “asked to assume that there [was] a clinically significant exposure to these chemicals among members of a group that is geographically defined as residing within Redlands.” (Italics added.) And Dr. Teitelbaum’s opinion that “risk of disease due to the toxins is spread over the whole exposed population” was qualified with the observation that “[t]he outcome of the exposure ... is determined by many factors including the dose, and the genetic makeup of the target individual.”
Moreover, regardless of how a particular medical monitoring class might be defined, a plaintiff must demonstrate that “the need for future monitoring is a reasonably certain consequence of [the] toxic exposure” (Potter, supra, 6 Cal.4th at p. 1009), i.e., that the plaintiff faces a “significant but not necessarily likely risk of serious disease” (id. at pp. 1008-1009). For the following reasons, we conclude plaintiffs have not placed in the record sufficient evidence to warrant the trial court’s concluding that they are likely to be able to make that demonstration with common proof.
Plaintiffs’ class definition refers to at least 12 different toxic substances, and plaintiffs contend that, as a consequence of defendants’ toxic dumping, each class member now requires special monitoring for numerous potential medical conditions. In linking their class definition to the toxic dumping and [1110]*1110water pollution evidence submitted in support of the certification motion, plaintiffs relied primarily on the testimony of two medical experts, Dr. James Dahlgren and Dr. Daniel Teitelbaum. We previously have held that reliable medical expert testimony may establish the reasonableness and necessity of medical monitoring. (Potter, supra, 6 Cal.4th at p. 1009.) “Expert medical opinion, however, does not always constitute substantial evidence . . . .” (Place v. Workmen’s Comp. App. Bd. (1970) 3 Cal.3d 372, 378 [90 Cal.Rptr. 424, 475 P.2d 656]; see, e.g., Mann v. Cracchiolo (1985) 38 Cal.3d 18, 40 [210 Cal.Rptr. 762, 694 P.2d 1134] [medical malpractice action]; Kerr v. Bock (1971) 5 Cal.3d 321, 324 [95 Cal.Rptr. 788, 486 P.2d 684] [res ipsa loquitur case]; Zemke v. Workmen’s Comp. App. Bd. (1968) 68 Cal.2d 794, 798 [69 Cal.Rptr. 88, 441 P.2d 928] [disability apportionment appeal].) No reason appears why in the medical monitoring context we should depart from our settled understanding that “[a]n expert’s opinion which rests upon guess, surmise or conjecture, rather than relevant, probative facts, cannot constitute substantial evidence” (Garza v. Workmen’s Comp. App. Bd (1970) 3 Cal.3d 312, 318, fn. 3 [90 Cal.Rptr. 355, 475 P.2d 451]).
Dr. Dahlgren testified in conclusionary fashion that “[a]ll persons who are at risk . . . should be in [a] monitoring program.” He testified generally that “chemical exposure in Redlands has resulted in an excess of certain cancers” and “[e]arly diagnosis and treatment for these cancers would improve the prospect of cure or long term remissions,” but he acknowledged that “[t]he precise dose of exposure experienced by each person cannot be determined exactly because of variability in the delivery of the water.” (Italics added.) He also conceded that “safe levels of exposure in such a setting are not known precisely . . . .” (Italics added.)
Dr. Teitelbaum opined that “any person who fulfills the class definition proposed in this case is at greater risk of developing cancer and other serious illness which is known by medical scientists and toxicologists to be associated with the chemicals at issue in this case.” But neither Dr. Dahlgren nor Dr. Teitelbaum categorically stated that mere qualification under the class definition demonstrates a need for medical monitoring irrespective of actual chemical dosages received.
We previously have noted that courts confronting medical monitoring claims may consider “the relative increase in the chance of onset of disease in the exposed plaintiff as a result of the exposure, when compared to (a) the plaintiff’s chances of developing the disease had he or she not been exposed, and (b) the chances of the members of the public at large of developing the disease” (Potter, supra, 6 Cal.4th at p. 1009). Indisputably, a member of the [1111]*1111public’s chances of developing any particular disease would be susceptible to common proof, but each individual plaintiff’s chances of developing that particular disease, had he or she not been exposed as alleged, may not be.
Taken as a whole, the medical expert testimony plaintiffs presented in support of their motion for class certification is too qualified, tentative and conclusionary to constitute substantial evidence that plaintiffs, by adopting a liability theory that makes actual dosages and variations in individual response irrelevant, will be able to prove causation and damages by common evidence. As the record stands, therefore, the causation and damages issues raised by plaintiffs’ claims must be counted among those that would be litigated individually, even if the matter were to proceed on a class basis. Especially when considered in light of the trial court’s finding that the class consists of an estimated 50,000 to 100,000 people, that conclusion fatally undermines the trial court’s predominance calculation.
In light of the foregoing, we conclude that the trial court’s predominance finding is not supported by the record. The questions respecting each individual class member’s right to recover that would remain following any class judgment appear so numerous and substantial as to render any efficiencies attainable through joint trial of common issues insufficient, as a matter of law, to make a class action certified on such a basis advantageous to the judicial process and the litigants. (Washington Mutual, supra, 24 Cal.4th at pp. 913-914.)
IV. Conclusion
Although the Court of Appeal erred to the extent it stated or implied that no action in which plaintiffs seek medical monitoring as a remedy may ever appropriately be certified for class treatment, we agree with the court that the trial court abused its discretion in granting the instant certification motion. (Linder, supra, 23 Cal.4th at pp. 435-436.)
Disposition
The judgment of the Court of Appeal is affirmed.
Kennard, J., concurred.