OPINION

COMPTON, Acting P.J.

Plaintiff Cinderella Jones instituted this action to recover damages for a cancerous condition she allegedly suffered as a result of ingestion of "Ortho-Novum SQ," a contraceptive pill developed and manufactured by defendant Ortho Pharmaceutical Corporation (Ortho). At the close of plaintiff's case-in-chief, the trial court granted defendant's motion for nonsuit on the ground that the evidence was insufficient to establish a reasonable causal connection between the development or aggravation of the disease and the use of the suspect drug. On this appeal, plaintiff contends that there was sufficient evidence present to warrant the case going to the jury. We disagree and thus affirm the judgment of nonsuit.

Viewing the record in the light most favorable to plaintiff, as we must, the evidence adduced at trial reveals the following chronology of events.

Plaintiff was born in May 1952. Shortly before her 17th birthday she experienced her first act of premarital sexual intercourse. Plaintiff subsequently married and had one child. Approximately six months before the birth of that child, plaintiff underwent a vaginal examination and was diagnosed as having venereal warts but an otherwise normal cervix.

By August 1970, plaintiff had divorced her first husband and remarried. Some three weeks prior to her second marriage, plaintiff began taking oral contraceptives when a physician at Riverside General Hospital prescribed Norinyl, a product not manufactured by defendant. Prior to receiving the drug, however, a Pap test was performed, which was reported as a class II smear. When the medication failed to alter her menstrual period, she returned to the hospital and received her first prescription of Ortho-Novum SQ. Plaintiff continued taking the drug for a period of six months through January 1971.

A second Pap smear performed in December 1970, was reported as a class I, but revealed that plaintiff had contracted a trichomoniasis infection. In May 1971, some three and one-half months after discontinuing the drug, a third Pap smear indicated a class III condition with "atypical squamous cells suggestive of mild dysplasia." A repeat Pap examination in June 1971, was again reported as a class III, showing "squamous dysplasia of cervix." A cervical biopsy performed one week later diagnosed plaintiff as suffering from "moderate squamous dysplasia." Another Pap smear taken in July 1971, produced a similar diagnosis.

In September 1971, plaintiff, then five months pregnant with her second child, underwent a cold conization procedure involving the removal of a small cone-shaped portion of her cervix. A subsequent biopsy of the cervical tissue revealed the presence of a precancerous condition known as carcinoma in situ. Thereafter, plaintiff had normal class I Pap smears, establishing that the cold conization had removed any trace of malignancy. One year later, however, plaintiff elected to have a vaginal hysterectomy by which her uterus was removed.

Ortho-Novum SQ manufactured by defendant until it was ordered off the market by the Food and Drug Administration in 1976, is termed a "sequential" birth control drug involving the administration of two different types of medication during a twenty-one-day period. The first tablet, containing an estrogenic compound known as mestranol, is generally taken for 14 consecutive days after which it is replaced by a second tablet composed of mestranol and a progestational compound, norethindrone. Ortho also manufactures a "combination" birth control drug, a mixture of mestranol and norethindrone, which is administered for 21 consecutive days of a 28-day cycle.

At trial, plaintiff contended that the ingestion of Ortho-Novum SQ was the cause of the carcinoma in situ or, in the alternative, that it caused an acceleration of the normal progression of the disease. Expert testimony was introduced to establish that carcinoma in situ of the cervix normally develops over a period of 86 months.

Plaintiff's condition, however, rapidly progressed from no dysplasia to carcinoma in situ within a 14-month period. Nevertheless, the medical experts called to testify at trial found it difficult, if not impossible, to establish a reasonable causal connection between ingestion of Ortho-Novum SQ and the development of plaintiff's disease.

Dr. Don Catlin, M.D., a specialist in pharmacology, testified that "there is a reasonable medical possibility," defined as less than a 50-50 chance, that the drug contributed to the development of the carcinoma in situ. His opinion was based in part on a review of 18 cases of carcinoma in situ which occurred during the premarketing clinical studies of Ortho-Novum SQ. Out of a total group of approximately 7,900 patients tested with the drug through February 1970, Dr. Catlin opined that there should have been only 2.04 cases of the disease instead of the 18 instances his study revealed.

Unlike Dr. Catlin, Dr. Michael Policar, M.D., a specialist in obstetrics and gynecology, was unwilling to state that the drug was a possible cause of plaintiff's carcinoma in situ. He testified "I think it [the birth control pill] may have been a contributing factor for the progression of [plaintiff's] lesion from dysplasia, whatever its starting point may have been to the end point of carcinoma in situ. I am specifically not implying that I think it caused the change from a normal to abnormal cervix."

In forming their opinions, both experts relied upon a study which revealed a statistically significant increase in the speed within which dysplasia develops into carcinoma in situ among women who use oral contraceptives. This increase, however, was observed only after five or six years of continuous use of such drugs. In relation to plaintiff's use of Ortho-Novum SQ for six months, Dr. Policar was unaware of any studies showing an acceleration of the disease process after such a short period of use.

Plaintiff's experts further testified that sexual intercourse at an early age, multiple partners, and the presence of venereal warts may also contribute to the onset of carcinoma in situ. Plaintiff denied, of course, that she fell into any of these categories.

Based upon the foregoing evidence, the trial court concluded that it was "beyond the capabilities of the jury to decide the issue of whether or not Ortho-Novum SQ caused plaintiff to suffer carcinoma in situ of the cervix, [or to determine] whether or not it accelerated the development of the dysplasia into a carcinoma in situ unless there was expert medical testimony on which the jury could reasonably base a finding of causation."

(1) A nonsuit may be granted only where, disregarding conflicting evidence on behalf of the defendant and giving to plaintiff's evidence all the value to which it is legally entitled, therein indulging in every legitimate inference which may be drawn from that evidence, the result is a determination that there is no evidence of sufficient substantiality to support a verdict in favor of the plaintiff. ( O'Keefe v. South End Rowing Club (1966) 64 Cal.2d 729, 733 [ 51 Cal.Rptr. 534, 414 P.2d 830, 16 A.L.R.3d 1]; Reynolds v. Willson (1958) 51 Cal.2d 94, 99 [ 331 P.2d 48].) Neither the appellate court nor the trial court may weigh the evidence nor consider the credibility of the witnesses. ( Lasry v. Lederman (1957) 147 Cal.App.2d 480, 488 [ 305 P.2d 663].)

(2) In determining whether a nonsuit was properly granted the reviewing court must resolve every conflict in testimony in favor of the plaintiff and at the same time indulge in every presumption and inference which could reasonably support the plaintiff's case. ( Salter v. Keller (1964) 224 Cal.App.2d 126, 128 [ 36 Cal.Rptr. 430].) (3) The rules governing the granting of a nonsuit, however, do not relieve the plaintiff of the burden of establishing the elements of his case. The plaintiff must therefore produce evidence which supports a logical inference in his favor and which does more than merely permit speculation or conjecture. ( Morgenroth v. Pacific Medical Center, Inc. (1976) 54 Cal.App.3d 521, 530, 531 [ 126 Cal.Rptr. 681].) If a plaintiff produces no substantial evidence of liability or proximate cause then the granting of a nonsuit is proper. ( Rufo v. N.B.C. Nat. Broadcasting Co. (1959) 166 Cal.App.2d 714, 720 [ 334 P.2d 16].)

(4a) In the instant case, the only evidence relating to the causal connection between Ortho-Novum SQ and plaintiff's precancerous condition is the highly conjectural and ambiguous testimony of Drs. Catlin and Policar that the ingestion of the drug may have had some effect on the development or progression of the disease. The question is whether such evidence meets the test of proximate cause. We believe it does not.

(5) The law is well settled that in a personal injury action causation must be proven within a reasonable medical probability based upon competent expert testimony. Mere possibility alone is insufficient to establish a prima facie case. ( Morgenroth v. Pacific Medical Center, Inc., supra, at p. 533; Johnston v. Brother (1961) 190 Cal.App.2d 464, 473 [ 12 Cal.Rptr. 23]; Pacific Employers Ins. Co. v. Industrial Acc. Com. (1960) 182 Cal.App.2d 162, 165 [ 5 Cal.Rptr. 738].) That there is a distinction between a reasonable medical "probability" and a medical "possibility" needs little discussion. There can be many possible "causes," indeed, an infinite number of circumstances which can produce an injury or disease. A possible cause only becomes "probable" when, in the absence of other reasonable causal explanations, it becomes more likely than not that the injury was a result of its action. This is the outer limit of inference upon which an issue may be submitted to the jury. (See Parker v. Employers Mutual Liability Ins. Co. of Wis. (Tex. 1969) 440 S.W.2d 43, 47.)

With cancer the question of causation is especially troublesome. Cancer, a generic term covering all malignant tumors, is one of the leading causes of death in the United States today. Under the present state of scientific knowledge, however, it is frequently difficult to determine the nature and cause of a particular cancerous growth. This point was emphasized by Dr. Catlin during trial when he testified, "[F]rom what we know about cancer and drugs it is very difficult to say that there is a clear cause and effect between almost anything or any drug or any toxin and carcinoma. . . ."

The fact that a determination of causation is difficult to establish cannot, however, provide a plaintiff with an excuse to dispense with the introduction of some reasonably reliable evidence proving this essential element of his case. (6) Although juries are normally permitted to decide issues of causation without guidance from experts, "the unknown and mysterious etiology of cancer" is beyond the experience of laymen and can only be explained through expert testimony. ( Parker v. Employers Mutual Liability Ins. Co. of Wis., supra, at p. 46.) Such testimony, however, can enable a plaintiff's action to go to the jury only if it establishes a reasonably probable causal connection between an act and a present injury.

On this issue we find persuasive the following discussion from Parker v. Employer's Mutual Liability Ins. Co. of Wis., supra: "[I]n the absence of factual circumstances of probability understandable to a jury there must be some scientific testimony that can be interpreted as an inference of hypothetical probability before we can allow a jury to speculate upon the rights of citizens. [¶] . . . If the experts cannot predict probability in these situations, it is difficult to see how courts can expect a jury of laymen to be able to do so. [¶] This requirement does in some instances place extraordinary burdens of proof on claimants. But once the theory of causation leaves the realm of lay knowledge for esoteric scientific theories, the scientific theory must be more than a possibility to the scientists who created it. For to the scientific mind, all things are possible. And with all things possible, citizens would have no reasoned protection from the speculations of courts and juries." ( Id. at p. 49.)

(4b) Our study of the record in the case at bench has convinced us that the only reasonable inference which can be drawn from the evidence is that the proximate cause of plaintiff's condition remains unknown and unproved. The testimony of plaintiff's experts was such that it is equally probable that the development of the carcinoma in situ was due to a cause for which defendant could not be liable. Here, specialists in both gynecology and pharmacology candidly admitted that the causal connection between the ingestion of Ortho-Novum SQ and plaintiff's condition was something less than a 50-50 possibility. On the issue of whether the drug accelerated the normal progression of the disease, the testimony was equally as uncertain and speculative. The experts found it difficult, if not impossible, to relate acceleration studies among women who took oral contraceptives for five or six years to someone like plaintiff who took the drug for only six months. Under the circumstances, we can only conclude that plaintiff did not establish a prima facie case and that the motion for nonsuit was properly granted.

(7) Relying upon Haft v. Lone Palm Hotel (1970) 3 Cal.3d 756 [ 91 Cal.Rptr. 745, 478 P.2d 465], plaintiff nonetheless contends that because defendant purportedly violated certain federal regulations in conducting its premarketing clinical studies of Ortho-Novum SQ, the burden of proof with respect to causation should have shifted to the defense. More specifically, she argues that "by not properly evaluating the question of whether or not its drug could cause carcinoma in situ, the defendant pharmaceutical company (along with the absence of any [other] published studies) has created a situation where plaintiff is unable to obtain expert opinion in terms of reasonable medical probability." We disagree.

In Haft, a father and son drowned in a motel pool where, in violation of statutory law, no lifeguard had been provided. There were no witnesses. Under those circumstances the Supreme Court held that the very violation of the statute — the failure to provide a lifeguard — made it impossible for the plaintiff to prove the cause of the drownings; hence the burden fell on the defendants to prove that the drownings would have occurred even if a lifeguard had been provided. In reaching this result the court reasoned that, "the shift of the burden of proof . . . may be said to rest on a policy judgment that when there is a substantial probability that a defendant's negligence was a cause of an accident and when the defendant's negligence makes it impossible, as a practical matter for plaintiff to prove `proximate causation' conclusively, it is more appropriate to hold the defendant liable than to deny an innocent plaintiff recovery, unless the defendant can prove that his negligence was not a cause of the injury." ( Id. at p. 774, fn. 19; italics added.)

Plaintiff maintains that federal regulations were violated which required defendant to design a study specifically for the purpose of evaluating whether or not Ortho-Novum SQ could cause carcinoma in situ of the cervix. Assuming arguendo that plaintiff succeeded in proving the regulatory violations, there would still be no justification for shifting the burden of proof to defendant to establish that the drug did not cause the development or aggravation of plaintiff's condition. Haft is simply inapposite to the case at bench.

The Haft rationale upon which plaintiff relies is in reality merely an extension of the doctrine of res ipsa loquitur. In Haft the physical cause of death was drowning. The instrumentality in which the drowning occurred, i.e., the pool water, was under the control of the defendant. Further there was a reasonable probability that the presence of a lifeguard would have prevented the drowning. Thus, the negligence being established, it was appropriate to presume causation.

Here plaintiff's case founders on the second step of the analysis. The first step, establishing the injury, was satisfied (here cancer and in Haft, drowning). The second step, the instrumentality which caused the injury and its control by the defendant (in Haft, the pool), is missing in plaintiff's case.

Plaintiff here would have us jump over step number two to step number three — the failure to comply with federal regulations (in Haft, the failure to have a lifeguard on duty) and from that point retroactively presume the existence of the second step.

Under these circumstances the presumption of causation would be tantamount to a presumption of the instrumentality which caused the injury. Such a quantum leap is justified by neither logic, legal precedent nor public policy.

We are unpersuaded by plaintiff's contention that defendant's failure to comply with federal regulations made it impossible for her to prove causation. Nor will we, as plaintiff inferentially suggests, presume that had defendant conducted the clinical studies plaintiff contends should have been done, those studies would have established that defendant's product causes cancer.

There is a limit to the number of presumptions in which the court will indulge solely for the purpose of assisting plaintiff in proving a case, especially when there is no evidentiary starting point from which those presumptions can flow.

The judgment is affirmed.

Beach, J., and Gates, J., concurred.

Appellant's petition for a hearing by the Supreme Court was denied March 27, 1985. Bird, C.J., was of the opinion that the petition should be granted.