This is a workmen’s compensation case in which Alton A. Parker brought suit against his employer’s insurer to recover for disability he alleged was caused by his employment. The trial court after a jury verdict rendered judgment in favor of the plaintiff, Parker, for total and permanent disability due to a cancer caused by exposure to radioactive materials in the course of his employment. The San Antonio Court of Civil Appeals has reversed and rendered judgment for the defendant on the grounds that there was no evidence of a causal connection between the plaintiff’s exposure to radiation and his subsequent development of cancer. 418 S.W.2d 570.
The question before this Court is whether there was evidence presented at trial that Parker’s cancer was caused by radiation received in the course of his em*45ployment. Parker, the petitioner here, contends that some evidence tending to prove causation was presented.
In April, 1961, the plaintiff went to work for his employer, a private company under contract to the United States Government at a government installation near San Antonio. He was initially employed as a material handler. His principal duty was to move material, some of which was radioactive, from one spot to another. He worked in that capacity until September, 1963, when he became a production operator. As a production operator, he assembled and disassembled nuclear weapons. The protective clothing he wore while working consisted of leaded gloves and a leaded apron similar to a carpenter’s apron, which apparently covered the chest and groin. On occasion while working as a production operator he also wore a film badge on his wrist beneath the leaded glove in order to record the amount of radiation which had penetrated the leaded glove. The badges were sent to a laboratory after being worn for a determination of the amount of exposure to radiation. An analysis of the badges worn by the plaintiff showed that he had been exposed to 36 millirems of radiation while wearing the badge.1 There is evidence in the record that supports a conclusion that since the badge was worn beneath protective clothing, it measured only part of the total amount of millirems to which the plaintiff was subjected. The evidence also shows that since material handlers were issued no measuring devices, it is impossible to tell the amount of radiation to which the plaintiff was subjected to while working as a material handler.
In April, 1965, Parker noticed a swelling on the left side of his neck. After consultation with his doctor, he saw Dr. James W. Nixon, Jr. On June 15, 1965, Dr. Nixon surgically removed from Parker’s neck what he diagnosed as a “metastatic carcinoma in the cervical lymph node.” Histological analysis suggested that the removed malignant cells were metastatic in character, reproducing in the lymph node a seminoma, or cancer of the testicle. However, no definite conclusion as to the site of the primary malignant tumor was reached.
Plaintiff, who has since died, brought suit in December, 1965, to recover for total and permanent disability which he alleged was caused by the radiation he received at his employment. Under Vernon’s Tex.Civ. Stat.Ann. Art. 8306, § 20(h) (1967), a workmen’s compensation provision, an employee is entitled to compensation if his disability resulted from a “diseased condition caused by exposure to x-rays or radioactive substances.”
In workmen’s compensation the injured employee need not prove fault, but he must prove that his injury arose out of his employment. This is essentially the same concept of causation in fact as that applied in a negligence suit, but it arises in a different context of liability. Thus the employer is responsible for a disease or injury, whether at fault or not, if the employment occasioned it. Our notions of justice and social policy have long required that there be this causal connection to assess liability for harm. The sine qua non test must be met: but for the conduct or condition, in this case the employment, would the plaintiff have suffered the harm?
The question now before us, then, is what evidence of causal connection between the employment and the plaintiff’s injury there must be to justify the judge submitting the case to the jury. With cancer this question of causation is especially troublesome. Cancer, a generic term covering all malignant tumors, is the second leading cause of death in the United *46States.2 The tumor, a neoplastic growth of cells which disturbs the normal relationship of tissues, is thought by some scientists to be induced by radiation and sunlight, parasites, and certain chemicals.3 Under the present state of scientific knowledge, however, it is difficult to determine the cause of particular cancerous growth. The determination of whether a cause is natural or artificial is especially difficult.4 Nevertheless, the fact that a determination of causation is difficult cannot provide a plaintiff with an excuse to dispense with the introduction of some evidence proving causation.
The common law has dealt with these difficult problems of causation by developing theories of causation upon which courts will submit an issue of fact to a jury. First, courts have allowed juries to decide causation where the general experience or common sense dictate that reasonable men know, or can anticipate, that an event is generally followed by another event. With the unknown and mysterious etiology of cancer, this is not the case we are here considering.
Secondly, courts present the jury with causation questions when there is a scientific generalization, a sharp categorical law, which theorizes that a result is always directly traceable back to a cause. Where, in other words, the harmful consequences provide a traceable chain of causation back to the act itself. This is the traditional use courts have made of expert testimony.5 Scientific generalizations of this sort provide the rationale for the “sequence of . events” test sometimes used when an injury coincides with a specific point of trauma, which we shall discuss later.
Thirdly, probabilities of causation articulated by scientific experts have been deemed sufficient to allow a plaintiff to proceed to the jury. For while a scientific training conceives of anything as possible, coincidence can be measured and generalizations similar to but not the same as uniform physical laws can be drawn from the probability of a result following a cause. In fact, the relationship between cause and its effect per se without theoretical explanation, can be nothing more than probable relationships between particulars. But this probability must, in equity and justice, be more than coincidence before there can be deemed sufficient proof for the plaintiff to go to the jury.
In Texas, expert medical testimony can enable a plaintiff’s action to go to the jury if the testimony is that there is a “reasonable probability” of a causal connection between an act and a present injury. Galveston H. & S. A. Ry. Co. v. Powers, 101 Tex. 161, 105 S.W. 491, 493 (1907).
In Insurance Co. of North America v. Myers, 411 S.W.2d 710 (Tex.Sup.1967) the substance of the doctor’s testimony was that it was only a possibility that the injury caused the death, not that it was or was not a reasonable medical probability. In denying recovery in that case we held: *47“Causal connection * * * must rest in reasonable probabilities; otherwise, the inference that such actually did occur can be no more than speculation or conjecture.” 411 S.W.2d 710, 713.
Although not in point to the case at bar, in Loper v. Andrews, 404 S.W.2d 300, 305, (Tex.Sup.1966), this Court held that a diagnostic entry in a hospital record is ad-missable under Article 3737(e) of the Texas Business Records Act, Vernon’s Tex.Civ.Stat.Ann. Art. 3737(e) (Supp.1966), only when the diagnosis of the recorded condition rests in reasonable medical certainty. This standard, that entries of medical opinions in hospital records must rest in reasonable medical certainty to be admissable, is even more stringent than the reasonable medical probability required to submit a causation issue to the jury.
Plaintiff by inference raises the question of whether there can be a logical distinction made between a reasonable medical “probability” and a medical “possibility.” We think that such a distinction can be made. There can be many possible “causes,” indeed, an infinite number of circumstances can cause an injury. But 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 can be submitted to the jury.
The existence of a “cause” is of course a matter of proof, and we now reach the question of whether there was any evidence that the radiation exposure did, in reasonable medical probability, cause the cancer Mr. Parker contracted.
The expert testimony was to the effect that the etiology of cancer is really unknown. The doctors further testified that the cancer “could have” been caused by radiation, but that there was no way to determine the cause. Dr. George C. Maney stated by deposition that it is possible for a person exposed to radiation over a long period of time to develop cancer, but that he could not state how much exposure would be required. Similarly, Dr. R. W. Bradley also stated by deposition in essence only that any radioactive material can conceivably cause cancer on prolonged exposure. Dr. Sidney Schiffer in summary testified only that anyone exposed to certain amounts of radiation has a higher than normal risk of developing malignant changes in the body tissues but that in this particular situation he “could not make a diagnosis of probability either way.” Dr. Vincent Collins testified that there was no way to determine what caused a particular cancer.
It is clear that there was no evidence at trial of causal connection between cancer and radiation in the expert testimony per se other than the possibility of such a connection. Indeed, the plaintiff’s efforts were directed towards only establishing a possibility, not a probability, when repeatedly he would ask of the experts: “Assuming a person was exposed over a long period of time in his work to radioactive and nuclear radiation material, is it possible for such a person to develop a cancerous condition in the body ?” Hence while this Court held in Myers that reasonable probability is determined by “the substance of the testimony of the expert witness and does not turn on semantics or on the use by the witness of any particular term or phrase,” the attempt here to elicit answers demonstrating possibility is in no way equivalent to testimony implicit with probability. 411 S.W.2d 710, 713.
We held in Otis Elevator Co. v. Lou Mae Wood, 436 S.W.2d 324 (Tex.Sup.1968), that it is not error for a plaintiff to attempt to show by questions asking what could be possible that the substance of an expert’s view is that a causal connection is a reasonable medical probability. But as we stated in Otis, this requires evidence that when the witness says something is *48possible he is using the phrase in the sense of what is probable. In this case the plaintiff at trial did not demonstrate that this is the proper interpretation of the expert testimony concerning possibility.
Petitioner argues that the testimony combined with the factual circumstances of this case creates a “reasonable medical probability” of causation. It is true that “reasonable medical probability” can be based upon the evidence as a whole. Galveston, H. & S. A. Ry. Co. v. Powers, 101 Tex. 161, 105 S.W. 491, 493 (1907); Atkinson v. United States Fidelity and Guaranty Co., 235 S.W.2d 509, 514 (Tex.Civ.App. — San Antonio 1950, writ ref’d n. r. e.). But this Court cannot agree that this situation is before us now. “Reasonable medical probability” from the whole evidence has arisen in some Texas cases which have allowed recovery for “traumatic cancer.” Trinity Universal Ins. Co. v. Walker, 203 S.W.2d 308 (Tex.Civ.App. —Austin, 1947, writ ref’d n. r. e.); Traders and General Ins. Co. v. Turner, 149 S.W.2d 593 (Tex.Civ.App. — Ft. Worth 1941, writ dism’d judgmt. cor.); see also Missouri-Kansas-Tex. R. Co. v. Evans, 151 Tex. 340, 250 S.W.2d 385 (1952). Contra, Jacoby v. Texas Employers Ins. Ass’n., 318 S.W.2d 921 (Tex.Civ.App. — San Antonio 1958, writ ref’d n. r. e.); Scott v. Liberty Mut. Ins. Co., 204 S.W.2d 16 (Tex.Civ.App. — Austin 1947, writ ref’d n. r. e.). In those few situations allowing recovery, where an employment trauma and a cancerous condition coincide at the same point of the body, some courts have held that it is reasonably probable that the cancer arises out of the course of employment. In these cases, despite medical science’s uncertainty as to the relationship between trauma and cancer,6 the trauma has been seen to be so related to the onset of cancer to allow a jury decision whether it was in fact the cause. In general, findings such as this occur when the trauma is an uncomplicated injury produced by a single mechanical force of which laymen can appreciate the consequences. Criteria have been developed by the medical profession to determine the probability of a causal relation between trauma and cancer.7 At a minimum these include: measurements of the authenticity and severity of the trauma; the origin of the cancer at the place of injury; and a “reasonable relationship” between the date of trauma, the appearance of the cancer, and the character or structure of the resulting growth.
There is no testimony or evidence other than coincidence supporting such a reasoned relationship between the radiation and cancer in this case. The extent of any radiation beyond the relatively safe dosage of 36 millirems is unknown. In fact, the probable origin of the cancer, the groin area, was part of the plaintiff’s body that was covered by the leaded apron and thus was in theory never exposed to over 36 millirems of radiation. Similarly the origin and structure of the tumor bears no probable relationship in the record to the radiation other than coincidence. Furthermore, the other causes such as natural radiation, virus, and infection have not been designated improbable by either the expert testimony or the circumstances of the plaintiff’s cancer.
In this case, then, there is simply no sequence of events strong enough to establish a probable causal connection. The Court of Civil Appeals is correct in holding that there is no evidence linking the radiation with the cancer other than the speculation and conjecture of laymen about an esoteric etiology science is only presently revealing.
Petitioner seeks us to restrict the Myers holding because among other reasons of the difficulty in getting medical experts *49to testify as to “probability.” We agree, as this Court stated in Myers, that particular words from the medical experts are not necessary to create a probability. 411 S.W.2d 710, 713. This is largely because medical experts define cause in terms of scientific exactitude, while the law is seeking an approximation of cause. But scientists first theorized as to the relationship between radiation and cancer, and in 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.
There is nothing in the record to suggest that the experts hypothesized the probable effect of the factual circumstances at bar on the complex etiological theory of radiation induced cancer. 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.8 And with all things possible, citizens would have no reasoned protection from the speculations of courts and juries.
Petitioner refers us to some jurisdictions which have allowed similar situations to go to the jury. In some states the doubts and burdens of proof are shifted from the claimant to the employer in workmen’s compensation cases. Scobey v. Southern Lumber Co., 218 Ark. 671, 238 S.W.2d 640, 243 S.W.2d 754 (1951); Royal Indem. Co. v. Land, 45 Ga.App. 293, 164 S.E. 492 (1932). Other states rely upon an implicit statutory presumption, as in Besner v. Walker Kidde Nuclear Laboratory, 24 A.D.2d 1045, 265 N.Y.S.2d 312, 313 (1965), where the court said: “The testimony of the medical experts is emphatic that there is really no ‘threshold’ of ‘safe’ dosage of radiation because at the present stage of scientific knowledge it cannot be ascertained exactly what effects radiation has on the human body. It is also admitted that each individual reacts differently to exposure to radiation. The award is supported by substantial evidence and by the presumptions (Workmen’s Compensation Law, §§ 3 subd. (2) and 47.)” The New York Statute referred to specifically speaks to disability caused by “exposure to ionizing rays.” N.Y. Workmen’s Compensation Law, Sec. 3(2), N.Y. Laws 1957, c. 906. New Jersey goes further and allows recovery for all diseases due to causes which are “characteristic” of a particular employment. N.J.Stat.Ann. 34:15-31. Ontario has perhaps advanced beyond any other jurisdiction in presuming causal connection with its statute creating a presumption of causal connection when any disease is contracted during an employment process listed with the disease in the statute. Ont. Rev.Stat. § 116(b). Some scholars have suggested doctrines of strict liability and res ipsa loquitur to deal with the possibility of injury induced by nuclear radiation.9
But with no statutory or common law rules creating an applicable policy otherwise this Court must require evidence of a probable causal relation between an injury and act before a case can go to a jury for a decision.
The judgment of the Court of Civil Appeals is affirmed.
*50Dissenting opinion by SMITH, J. REAVLEY and McGEE, JJ., not sitting.. A millirem is one thousandths of a rem. A rem is a rad, or unit of radiation absorbed dose, multiplied by an index of relative biological effect.
. Adelson, “Injury and Cancer,” in Physician in the Courtroom 9 (Schroeder ed:, 1954).
. Brooke, In the Wake of Trauma, 420 (1957); Russell and Clark, “Medico-Legal Considerations of Trauma and other External Influences in Relationship to Cancer,” 6 Vand.L.Rev. 868, 873-75, 877-87 (1953) ; Warren, “Criteria Required to Prove Causation of Occupational or Traumatic Tumors,” 10 Chi.L.Rev. 313, 316-17 (1943).
. “The relationship between radiation exposure and subsequent pathological conditions is perhaps the most difficult and urgent causation problem facing law and medicine today.” Curren, Law and Medicine, 27-118, p. 115 (1960).
. “The expert is in effect not telling of facts at all, but of uniform physical rules, natural laws, or general principles, which the jury must apply to the facts.” Hand, “Historical and Practical Considerations Regarding Expert Testimony,” 15 Harv. L.R. 40, 50 (1901).
. Moritz, Pathology of Trauma 116 (1954).
. Ewing, Neoplastic Diseases, 108 (4th ed. 1940); Also Moritz, Pathology of Trauma (1954); Willis, Pathology of Tumors (1st ed. 1948).
. “No trained scientist, medical or other, would have the effrontery to state today that space travel is impossible, that the discovery of a cure for leukemia or other forms of cancer is impossible, or that any other speculation is dogmatically beyond the realm of knowledge or accomplishment.” Matter of Miller v. National Cabinet, 8 N.Y.2d 277, 204 N.Y.S.2d 129, 134, 168 N.E.2d 811, 815 (1960).
. Cavers, “Improving Financial Protection of the Public Against the Hazards of Nuclear Power,” 77 Harv.L.R. 644, 670 (1964); Stason, Estep and Pierce, Atoms and the Law, 421-532 (1959).