Russell E. Harford, Jr., a long time insulation installer for defendant Bath Iron Works Corp. (BIW), had to quit work because of shortness of breath. Suit is brought on his behalf by the Director, Office of Workers’ Compensation Programs, U.S. Department of Labor, under the Longshore and Harbor Workers’ Compensation Act (LHWCA), 33 U.S.C. § 901 et seq., for work-related disability due to “Asbestosis and related diseases.” It is undisputed that Harford developed lung cancer, and further that his smoking two plus packs of cigarettes a day for 32 years was a basic cause. It is also undisputed that, though he did not work with asbestos, he was exposed to the dust from neighboring workplaces. He seeks to bring in this exposure as a contributor to his cancer in order to charge the employer and its insurer.
Trial was had before an ALJ on letters and depositions of medical experts. In his decision, the ALJ stated that the asbestos did not have to be the “sole cause” of the cancer for claimant to recover. Rather, the entire disability would be compensable if the asbestos contributed to, combined with, or aggravated it.1 In short, the ultimate question was whether there was a “causal relationship between Claimant’s employment and his lung cancer.” The first issue, however, was whether the employer had met the statutory presumption in claimant’s favor. Title 33 U.S.C. § 920(a) presumes, “in the absence of substantial evidence to the contrary—(a) That the claim comes within the provisions of [the Act].” After extensive review and discussion of the evidence, the ALJ found that the presumption was rebutted, which caused it to “fall”. See Sprague v. Director, O.W.C.P., 688 F.2d 862, 865 (1st Cir.1982). Weighing the evidence without the presumption, he found there was no causal relationship between claimant’s employment and his cancer. It is to be stressed that this was not for lack of proof by the claimant, but by belief of affirmative evidence submitted by the employer. Claimant did not, he found, contract asbestosis, and asbestos without asbestosis did not cause or contribute to the cancer.
The Benefits Review Board reversed the ALJ’s finding that the presumption in claimant’s favor had been rebutted, thereby ending the defense. We reverse.
Decision of the Benefits Review Board
With reference to the presumption, the Board stated, “[The] employer’s burden on rebuttal [is] to present specific and comprehensive evidence sufficient to sever the causal connection between the injury and the employment.” It added, “The unequivocal testimony of a physician that no relationship exists ... is sufficient to rebut the presumption.” This positive language it recast as “unequivocally severs.” “Unequivocal,” on a search of dictionaries, universally means “not doubtful,” or the like, which we take to mean certainty. That the Board so intended is inescapably confirmed by its response to employer’s expert, Dr. Cadman’s unwillingness *675to be absolutely certain. Because Dr. Cad-man, in the Board’s words, conceded that “asbestos may have contributed to claimant’s lung cancer and that he could not exclude that exposure as having contributed” to it, the Board thought his testimony insufficient to rebut the presumption.
As we have previously held, the presumption is overcome with substantial evidence of non-causation. See Sprague, 688 F.2d at 865. Substantial evidence is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Id. (quotations and citation omitted). This means “reasonable probabilities.” Cf. DaSilva v. American Brands, Inc., 845 F.2d 356, 361 (1st Cir.1988); Bath Iron Works Corp. v. Director, O.W.C.P., 109 F.3d 53, 56 (1st Cir.1997); Oberlander’s Case, 348 Mass. 1, 7, 200 N.E.2d 268 (1964) (Workmen’s Compensation). Dr. Cadman’s medical opinion was found insufficient by the Board, however, because he could not exclude possibilities—a typical expert opinion. This put an impossible burden on the employer.
We have, therefore, two questions. Did the employer submit substantial evidence, when properly defined? (A question of law for the court, not dependent on credibility. See Sprague, 688 F.2d at 865; CNA Ins. Co. v. Legrow, 935 F.2d 430, 433-34 (1st Cir.1991)). At the same time, we may ask whether the ALJ was warranted in his substantive findings. In this connection we note 33 U.S.C. § 921(b)(3),
The findings of fact in the decision under review by the Board shall be conclusive if supported by substantial evidence in the record considered as a whole.
This means, obviously, that the ALJ’s choice of inferences is to be respected. See Sprague, 688 F.2d at 866 (“In reviewing for substantial evidence it is immaterial that the facts permit diverse inferences as long as those drawn by the ALJ are supported by evidence.”).
The ALJ’s Decision
The ALJ’s conclusions are backed by an extensive discussion of the several experts. One or more of BIW’s experts testified that current medical evidence indicates that a finding of interstitial fibrosis is- necessary to reach a diagnosis of asbestosis. There was testimony that tissue samples and x-ray analyses produced no evidence that Harford had asbestosis; that he did not have fibrosis at the time of his surgery, and that the alveolar damage found in a biopsy some months later was caused by radiation and • chemotherapy rather than by asbestos exposure.
Two experts testified on causation of the cancer. Dr. Cadman’s conclusion was,
[C]urrent evidence evaluating the association of asbestos exposure, lung fibrosis and lung cancer strongly suggest [sic] that the excess lung cancer attributable to asbestos is associated with fibrosis. Therefore, lung cancer in the absence of pulmonary fibrosis is most likely lung cancer which developed either from the effects of smoking alone or it arose unrelated to any known carcinogen as occurs in the non-smoker____ I believe that in the absence of fibrosis, that [Mr. Harford’s] lung cancer was most likely the result of prior smoking history, (emphasis added).
The ALJ concluded,
Based on the record medical evidence, I determine that the employer/carriers have produced specific objective clinical evidence and soundly reasoned medical opinions sufficient to sever the causal nexus and have thereby rebutted the Section 20 presumption.
Since the ALJ had stated earlier that cause included contribution, we find that the ALJ was warrantably covering both in this finding. Dr. Cadman reasonably could be found as of the opinion that diffuse interstitial fibrosis (and therefore asbestosis) is generally present when asbestos exposure is a contributing cause to lung cancer. Viz., the probability is against asbestos," in the absence of asbestosis. In other words, asbestos exposure had had no belated, contributory, effect, as well as no original effect. We believe, following Dr. Cadman’s reference to “excess” lung cancer, particularly when there had been negative evidence covering the period between surgery and the subsequent biopsy, that his last quoted sentence is not to be read *676as limiting his opinion to the original cancer. In short, Dr. Cadman’s opinion is substantial evidence of non-causation, sufficient both to rebut the presumption and to support a finding for BIW.
Next?
First, an observation. Harford’s claim was filed on March 20, 1989, just short of nine years ago. The ALJ’s unfavorable decision on the merits was filed on June 10,1991; the Board’s remand for a hearing on damages on November 23, 1993. BIWs petition for review of the ALJ’s damage finding was filed July 7,1994. This finding was affirmed automatically under the statute, Omnibus Consolidated Rescissions and Appropriations Act of 1996, Pub.L. No. 104-134, § 101(d), 110 Stat. 1321, 1321-219 (1996), because of the Board’s not having passed on the petition before September 12, 1996. We believe the time has come to resolve matters. If we remand the case to the Board to review the ALJ’s decision on the merits, it will be bound by his findings in the absence of error. See 33 U.S.C. § 921(b)(3). Why the unnecessary steps? The short answer is that we should take over finally now. And, indeed, we have already reviewed the ALJ’s detailed and careful opinion. Substantial evidence supports the ALJ’s conclusion that there was no causal relationship. We find no error.
The decision of the Benefits Review Board is reversed, and the denial of claim of Russell E. Harford, Jr. is reinstated.
. We note that the Board, similarly to the ALJ, viewed the issue as whether "claimant's exposure to asbestos did ... cause or contribute to his lung cancer.” The parties share this conception. Thus, we understand "aggravate” to be essentially synonymous with “contribute,” and to refer to the cancer. No claim has been made for aggravation of associated symptoms. Cf. Gardner v. Director, O.W.C.P., 640 F.2d 1385, 1389 (1st Cir.1981).