The defendant, Diane Johnson (claimant), appeals from an order of the Superior Court (Contas, J., approving the report of a Master, R. Peter Shapiro, Esq.) denying workers’ compensation benefits for her husband’s fatal heart attack. RSA 281:2, V (Supp. 1986). We reverse and remand.
Jeremiah Johnson, age forty-seven, went to work for Bartlett Tree Experts Company (Bartlett) as a temporary groundman on July 23, 1984. He had worked for Bartlett in that capacity previously, but in the year prior to his death, Johnson had been employed at a desk job for eight months and, for several days, as a mechanic. This was his first day back on the job with Bartlett.
*705Johnson and John Sprano, a Bartlett foreman, began work at 7:30 a.m., pruning trees near power lines in Etna for the Granite State Electric Company. Johnson’s job involved operating a “chipping” machine, collecting fallen branches, pulling down tree limbs, and, when necessary, felling trees.
The day was hot and humid, and the temperature rose above 95-degrees in the afternoon. Granite State Electric Company called in all its outside employees that afternoon because of the heat, but Johnson and Sprano, as employees of Bartlett, continued their work. At about 3:00 p.m. they stopped for a water break, and Johnson vomited about five minutes later. Although he complained about the heat, the men continued to work until 3:45 p.m.
After work, they started for home in Sprano’s pick-up truck. Sprano stopped at a grocery store, and Johnson remained outside. When Sprano returned, Johnson was outside the truck vomiting. Johnson said he would be “all right in a minute,” and they resumed their trip home. A short time later, Johnson became unresponsive to Sprano’s questions, and when Sprano raised his voice, Johnson suddenly collapsed.
Sprano stopped to summon help, and an ambulance arrived several minutes later. Efforts to revive Johnson were unsuccessful, and, at 5:45 p.m., he was pronounced dead at Mary Hitchcock Hospital in Hanover. The cause of death was cardiac arrest due to ventricular fibrillation. Johnson had no prior history of heart disease, and no autopsy was performed.
Mrs. Johnson seasonably filed for workers’ compensation benefits. A hearings officer of the New Hampshire Department of Labor awarded benefits based on his determination that “Johnson did suffer an accidental injury arising out of and in the course of his employment at Bartlett Tree Experts Company.” The plaintiffs, Bartlett and its workers’ compensation carrier, Aetna Life and Casualty Company, appealed the decision to the superior court. After a trial de novo, the master found that the claimant had failed to “establish to a reasonable degree of medical certainty that the deceased’s employment substantially contributed” to his fatal heart attack. The superior court, approving the master’s report, denied coverage, and the claimant appealed.
On appeal, the claimant argues that: (1) there was sufficient evidence to prove medical causation; (2) the master erred in his reliance on the testimony of the plaintiffs’ expert; and (3) the master erred in his application of the substantial contribution test to determine legal and medical causation in this case. The plaintiffs, *706on the other hand, argue that the master’s report is supported by the evidence and free of legal error.
We begin with a brief consideration of the claimant’s argument that the testimony of her expert, Dr. Yanofsky, was sufficient to establish a medical causal connection between her husband’s employment and his heart attack. Dr. Yanofsky, the director of Advanced Cardiac Life Programs at Dartmouth-Hitchcock Medical Center, testified, inter alia, that the combined effect of “[t]he heat, humidity, and [the deceased’s] exertion” contributed to his heart attack. While we agree that this testimony would have been sufficient to establish medical causation if the master had accepted it, see, e.g., Carpino v. Treasure Chest Restaurant, 106 A.D.2d 782, 483 N.Y.S.2d 817, mod. on other grounds, 65 N.Y.2d 782, 492 N.Y.S.2d 948, 482 N.E.2d 566 (1984) (expert testimony that chef’s employment, which required long hours and heavy lifting in a hot kitchen, was causally connected to his heart attack sufficient to sustain award of benefits); Couture v. Mammoth Groceries, Inc., 116 N.H. 181, 183, 355 A.2d 421, 422-23 (1976) (expert testimony that meatcutter’s heart attack was probably caused by lifting heavy beef quarters sufficient to establish medical causation), the master had conflicting expert testimony before him, and a trier of fact is free to accept or reject an expert’s testimony, in whole or in part, when faced with conflicting expert testimony. Morrill v. Tilney, 128 N.H. 773, 778, 519 A.2d 293, 297 (1986); cf. Town of Hudson v. Wynott, 128 N.H. 478, 485-86, 522 A.2d 974, 978 (1986) (where issue is within the province of medical experts, master required to identify the consideration which impelled decision to disregard uncontradicted medical testimony). Thus, the testimony of the claimant’s expert, standing alone, provides no basis to reverse the decision of the superior court.
We turn next to the argument that the master’s reliance on the testimony of the plaintiffs’ expert, Dr. Ross, requires reversal because his testimony was based on possibilities rather than probabilities. As a preliminary matter, we note that the cases relied on by the claimant are distinguishable from this case because the parties seeking to introduce the expert testimony in those cases bore the burden of proof. In Emerson v. Company, 87 N.H. 108, 112, 174 A. 779, 782 (1934), for example, the claimant’s expert admitted that he could “say more about the possibilities than . . . the probabilities.” This court held that this testimony should not have been received without corroborative evidence because “it related to mere *707possibilities, or to things not more probable than otherwise, [and] it would afford no basis for an assessment of damages.” Id. The court’s holding in Emerson was clearly based on the allocation of the burden of proof to the claimant, whose own witness’s testimony was at issue. In this case, the claimant also had the burden to prove that she was entitled to workers’ compensation benefits, but it is the testimony of the plaintiffs’ expert that is at issue. However, we need not rest our decision on that ground because the claimant’s argument mischaracterizes Dr. Ross’s testimony. The substance of Dr. Ross’s testimony was that he believed there were three possible medical explanations for the deceased’s cardiac arrest: (a) myocardial infarction; (b) sudden cardiac death syndrome; and (c) perforation of an ulcer. In Dr. Ross’s opinion, based on reasonable medical probabilities, the heat, the humidity, and the deceased’s exertion were not significant factors in causing death because his cardiac arrest would have occurred “irrespective of [his] work environment.” Therefore, the claimant’s argument is without merit.
Another aspect of the master’s reliance on Dr. Ross’s testimony is noteworthy. Relying on Dr. Ross’s testimony, the master discounted heat as a factor in causing Johnson’s heart attack because “heat is more detrimental . . . when fluids are not taken and when the individual becomes dehydrated.” Dehydration occurs when a person “loses too much water, as through excessive vomiting.” 1 J. Schmidt, Attorney’s Dictionary of Medicine, D-25 (1986). Although the deceased did take in some fluids at lunch, he worked outside in 95-degree heat on the afternoon of his death, and at 3:00 p.m., when he stopped for a water break, he vomited. He continued working, and approximately an hour later, shortly before his heart attack, he vomited again. We need not consider whether the master erred in finding that the deceased was not dehydrated in these circumstances because we reverse the superior court’s decision on other grounds. We do, however, remind the trial courts that although the weight to be accorded to expert testimony is generally a determination for the trier of fact, see Morrill v. Tilney supra, the bases of expert testimony must be carefully considered. “[T]he opinions expressed by any expert are only of value insofar as they are based upon factual assumptions which are fairly supported in the record.” Johnson v. Califano, 434 F. Supp. 302, 309 (D. Md. 1977); see Calhoun v. Honda Motor Co., Ltd., 738 F.2d 126, 129 (6th Cir. 1984); see also N.H. R. Ev. 703; 11 J. Moore, Moore’s Federal Practice § 703.10 (2d ed. 1985) (Fed. R. Evid. 703, from which N.H. R. Ev. 703 is derived, assumes that expert testimony must have a factual basis).
*708The claimant’s next argument is more persuasive. The claimant contends that the master erred in applying the substantial contribution test in this case. Generally, a claimant in a workers’ compensation case “must prove two required elements: that the [employee’s] injury or death was accidental; and that the injury or death was caused by his employment.” New Hampshire Supply Co. v. Steinberg, 119 N.H. 223, 226, 400 A.2d 1163, 1165 (1979), appeal after remand, 121 N.H. 506, 433 A.2d 1247 (1981); RSA 281:2, V (Supp. 1986). The plaintiffs concede that the deceased’s heart attack was accidental in the sense that it was unexpected. Thus, the critical issue at trial was causation.
In Steinberg, this court adopted Professor Larson’s suggested analysis for the difficult problem of determining the compensability of heart attacks. See Larson, The “Heart Cases” in Workmans’ Compensation: An Analysis and Suggested Solution, 65 Mich. L. Rev. 441 (1967). A claimant in a heart case must prove both legal and medical causation between work-related stress or exertion and the heart attack. The appropriate test to determine legal causation depends on the employee’s prior health:
“Where there is a prior weakness in the form of a previously weakened or diseased heart, then the employment must contribute something substantial to the heart attack. That is, the employment-connected stress or strain must be greater than is encountered in normal nonemployment life. Thus, heart attacks that actually result from work related stress are distinguished from those that occur at work merely as a result of natural physiological process. If there is no prior weakness or disease of the heart, any [work-related] exertion connected with the heart as a matter of medical fact is adequate to satisfy the legal test of causation so as to make the injury or death compensable.”
Steinberg, supra at 231, 400 A.2d 1168 (citations omitted) (emphasis in original).
Under Steinberg, the substantial contribution test of legal causation is appropriate only when the existence of a pre-existing weakness or disease is proved by the party alleging it. IB A. Larson, Workmen’s Compensation Law § 38.83, at 7-294 (1987). When the existence of a prior heart disease or weakness is established, the claimant satisfies the substantial contribution test of legal causation by proving that the work-related stress or exertion which allegedly caused the heart attack was greater than *709that which a person normally encounters in non-employment life. Steinberg, 119 N.H. at 231, 400 A.2d at 1168; Guidry v. Sline Indus. Painters, 418 So. 2d 626, 633 (La. 1982); see, e.g., 1B A. Larson, supra at 7-280. Medical testimony is not necessary on this issue. See 1B A. Larson, supra at 7-280-81.
Regardless of which test of legal causation is appropriate in a given case, the test of medical causation remains the same: did the work-related stress or exertion probably cause or contribute to the employee’s heart attack as a matter of medical fact? (“Probably” in this context refers to the preponderance of the evidence standard.) Steinberg, 119 N.H. at 231, 400 A.2d at 1109; Price River Co. v. Industrial Com’n, 731 P.2d 1079, 1082 (Utah 1986).
In this case, the plaintiffs argued that the substantial contribution test was the appropriate test of legal causation because the deceased allegedly suffered from heart disease prior to his death. Therefore, the plaintiffs had the burden to establish the existence of that prior heart condition. 1B A. Larson, supra at 7-294. Although the deceased had no prior symptoms or medical history of heart disease, the master determined that the substantial contribution test of legal causation was appropriate based on the statistical assumption of the experts that the deceased had prior heart disease because he would not otherwise have had a heart attack. Compare Lamoreux v. W.C.A.B. (Celotex Corp.), 92 Pa. Commw. 1, 497 A.2d 1388 (1985), with 1B A. Larson, supra at 7-291-94.
The master could not reasonably do this in light of Dr. Ross’s repeated testimony that he did not know the cause of Johnson’s death. Thus, there was no evidentiary predicate on the record in this case to require the “substantial contribution” test. It necessarily follows that the master’s determination that compensation should be denied because “the [claimant] did not establish to a reasonable degree of medical certainty that Jeremiah Johnson’s employment substantially contributed to his death” was error. (Emphasis added.) This created an erroneous combination of the issues of legal and medical causation and improperly raised claimant’s burden of proof on the issue of medical causation. Therefore, we reverse the decision of the superior court, and remand the case for a new trial.
Reversed and remanded.
All concurred.