OPINION
PER CURIAM.Following multiple hearings and agency-decisions, the Benefits Review Board of the Department of Labor awarded black lung benefits to Clyde Breeding on February 8, 2002. On appeal, Colley & Colley Coal Company contends that the Board erred in concluding that Breeding had worked at least 15 years in coal mine employment, thus entitling him to a regulatory presumption of total disability due to pneumoconiosis, and that Colley & Colley’s medical evidence failed to rebut that presumption. For the reasons that follow, we affirm the Board’s conclusion that Breeding worked as a coal miner for at least 15 years, but we remand on its conclusion that Colley & Colley’s medical evidence did not rebut the presumption that Breeding’s disability was due to pneumoconiosis.
I
Clyde Breeding worked as a coal miner for several different coal companies from 1941 to 1970, last working in that capacity for Colley & Colley Coal Company. Thereafter, he worked another 13 years as a State mine inspector. During his entire employment — a period of more than 40 years — Breeding smoked a pack of cigarettes a day until he quit in 1983.
On September 29, 1980, Breeding filed a claim for benefits under the Black Lung Benefits Act, and, following a hearing, the Administrative Law Judge (“ALJ”) awarded Breeding benefits.
Following an appeal of that award, the Benefits Review Board (“Board”) remanded the case to the ALJ to reconsider whether Colley & Colley’s medical evidence rebutted the presumption under 20 C.F.R. § 718.305 of total disability due to pneumoconiosis, which the Board found applicable based in part on its inclusion of Breeding’s government mine inspection work as qualifying employment for purposes of the Act. Colley & Colley filed a motion with the Board for reconsideration.
While the motion for reconsideration was pending, on March 25, 1992, Breeding died. His wife, Delta Breeding, thereafter filed a claim for survivors’ benefits. After an initial determination of eligibility, Delta Breeding’s file was forwarded for association with Breeding’s original claim. In addition, Delta Breeding continued her husband’s original claim.
The Board granted Colley & Colley’s motion for reconsideration and, after rehearing the case en banc, concluded that qualifying employment did not include Breeding’s years as a government mine inspector. The Board remanded the case to the ALJ for recalculation of Breeding’s years of coal mine employment.
On remand, the ALJ (a different one) denied benefits. In his decision and order dated December 12, 1995, the ALJ found the evidence insufficient both to establish pneumoconiosis and to invoke the § 718.305 presumption that Breeding’s total disability was due to pneumoconiosis. Considering whether Breeding worked at least 15 years to invoke the presumption, the ALJ
employed a methodology whereby he adopted the highest quarterly wage earned during specific periods as representing full-time employment for such periods and assigning a percentage of full-time work for quarters where [Breeding] earned less than this amount.
Employing this method, the ALJ found that between 1941 and 1963, Breeding had worked 4.46 years in coal mine employment. Adding that figure to the 8 years from 1963-1970 conceded by Colley & Col*565ley, the ALJ credited Breeding with 12.46 years of coal mine employment. With this finding, Breeding was not entitled to the presumption under § 718.305.
On appeal, the Board affirmed the finding that Breeding had 12.46 years of qualifying coal mine employment, but it remanded to the ALJ for reconsideration of the medical evidence regarding pneumoconiosis.
On remand, the ALJ (again a different one), relying in part on evidence contained in the file in the claim of Breeding’s widow, awarded benefits.
On appeal, the Board again vacated the award and remanded for consideration of Breeding’s claim based only on the record in his case or, alternatively, for consolidation of the two claims and further development of the widow’s record, which had been improperly curtailed.
Again on remand, the ALJ consolidated the two claims and remanded the case to the District Director for further development of the record in the consolidated case. The District Director, upon accepting additional evidence, recalculated Breeding’s length of employment at 16.75 years and awarded benefits. The new evidence included the fact that Breeding’s employment at Wright’s Super Market was actually employment in Wright’s coal mine operation, not at the supermarket itself.
Following the District Director’s new findings, the case was transferred back to the ALJ, who awarded benefits on October 30, 2000. The ALJ (yet again a different one) found that Breeding was entitled to the § 718.305 presumption and that Colley & Colley had failed to rebut the presumption. As to Breeding’s coal mine work history, the ALJ followed the District Director’s method:
[T]he method used to arrive at the figure of 16.75 years of coal mine employment involved identifying the [Bureau of Labor Statistics] average daily wage earned by miners in a particular year and multiplying that figure by 125 to arrive at an “Earning Standard” for each year that Mr. Breeding engaged in coal mine employment. If Mr. Breeding’s actual annual wages, as recorded by the Social Security Administration, exceeded the Earnings Standard for that year, he was credited with a year of coal mine employment.
Using this method, the ALJ found that Breeding had worked 16.75 years in coal mine employment from 1941 through 1970 and thus was entitled to the § 718.305 presumption that his disability was due to pneumoconiosis.
In his analysis of the medical rebuttal evidence offered by Colley & Colley, the ALJ discounted the reports of four physicians (Drs. Castle, Dahhan, Hansbarger, and Tomashefski), who gave their opinion that Breeding’s disability was not caused by coal mine employment but rather by smoking. Because these physicians discussed the medical, not legal, definition of pneumoconiosis, the ALJ rejected their reliability and relied instead on the reports of five other physicians (Drs. Buddington, Garzón, Kanwal, O’Neill, and Schmidt), who attributed Breeding’s disability either to a chronic respiratory impairment of which pneumoconiosis was a contributing factor or to an obstructive airways disease with a restrictive component, which can be a manifestation of pneumoconiosis.
The Board affirmed the ALJ’s award of benefits in a decision and order dated February 8, 2002. The Board agreed that Breeding worked 16.75 years in coal mine employment and was therefore entitled to the § 718.305 presumption and that the medical evidence submitted by Colley & Colley failed to rebut the presumption. *566To challenge both of these conclusions, Colley & Colley filed this Petition for Review.
We review the ALJ’s findings as affirmed by the Benefits Review Board to determine whether they are in accordance with the law and supported by substantial evidence. Island Creek Coal Co. v. Compton, 211 F.3d 203, 207 (4th Cir.2000). In our review, we will “not reweigh the evidence or substitute our views for those of the ALJ,” Lane v. Union Carbide Corp., 105 F.3d 166, 170 (4th Cir.1997), but we must address “whether all of the relevant evidence has been analyzed and whether the ALJ has sufficiently explained his rationale in crediting certain evidence,” Milburn Colliery Co. v. Hicks, 138 F.3d 524, 528 (4th Cir.1998). As to medical evidence in particular, we recognize that “it is the province of the ALJ to evaluate the physicians’ opinions,” but nonetheless the ALJ “must examine the reasoning employed in a medical opinion in light of the objective material supporting that opinion, and also must take into account any contrary results or diagnoses.” Island Creek Coal Co., 211 F.3d at 211. We review decisions of law de novo. Armco, Inc. v. Martin, 277 F.3d 468, 473 (4th Cir.2002).
II
Colley & Colley first contends that the Board’s conclusion that Breeding worked for 16.75 years in coal mine employment, entitling him to the presumption under 20 C.F.R. § 718.305, was erroneous. It argues that the ALJ’s 1995 calculation of 12.46 years, which had been affirmed by the Board, became the law of the case and therefore precluded subsequent departure from that calculation and its underlying formula. Breeding relied on the § 718.305 presumption to establish the necessary element of his claim that his total disability was due to pneumoconiosis. See Island Creek Coal, 211 F.3d at 207; 20 C.F.R. §§ 718.201-.204.
In recalculating the years of employment in his 2000 decision and order, the ALJ stated that the previous method was not law of the case because “it involves a question of fact and not one of law.” Noting the additional evidence submitted by Breeding’s widow, as well as several other periods of coal mine employment that were reflected on Breeding’s Social Security Itemized Statement of Earnings not preciously included, the ALJ stated that the calculation of 12.46 years of coal mine employment was a finding “based, at least in part, on a mistake of fact.” The ALJ conceded that if he adhered to the underlying method, Breeding’s years of coal mine employment would amount to 14.18 years, but because the ALJ considered the approach used by the District Director to be “fairer” than that used in 1995, he concluded that 16.75 years was the proper amount of qualifying employment. The Board affirmed.
We conclude that the Board did not err in this regard. In 1999, it permitted further evidentiary development of the record relevant to both the miner’s claim and the widow’s claim if the ALJ chose to consolidate the two, and pursuant to that grant, new evidence regarding Breeding’s coal mine work history was submitted to the District Director. By permitting new evidence and new findings, the Board necessarily reopened the previous ruling on work history because the evidence, as expanded, no longer supported the finding of 12.46 years. Accordingly, that finding, along with the facts supporting it, was no longer binding on the ALJ. The law-of-the-case doctrine is not so restrictive that it binds a judicial officer, entitled to take new evidence and make new findings, to an earlier finding — even if it was a mixed question of law and fact — particularly when the preexisting evidence no longer supported the finding asserted to be the *567law of the case. See Sejman v. Warner-Lambert Co., 845 F.2d 66, 69 (4th Cir.1988) (stating that the judicial doctrine does not operate when “a subsequent trial produces substantially different evidence”); Smith v. North Carolina, 528 F.2d 807, 810 (4th Cir.1975) (stating that the district court was not bound by law of the case when new evidence established a conclusion previously found erroneous on appeal).
Furthermore, we find no error in the particular method employed by the ALJ. As the Board found, Colley & Colley “has not identified any factual error produced by the application of this method or explained how it has suffered actual prejudice that would be rectified on remand.” Because the Black Lung Benefits Act, at the time relevant to this appeal, permitted any evidence to establish “regular employment,” see 20 C.F.R. § 718.301(a) (2000), and calculated a year of “regular employment” as a minimum of 125 work days, see Armco, 277 F.3d at 472-75, we conclude that the ALJ’s method, ultimately finding 16.75 years of qualifying employment, was not error.
Accordingly, we deny Colley & Colley’s petition for review on this issue.
Ill
Colley & Colley also contends that the ALJ erred in his weighing of the medical evidence submitted to rebut the § 718.305 presumption. It argues that the ALJ improperly discounted the opinions of certain physicians based on “the judge’s determination that these physicians relied upon a ‘medical’ not the ‘legal’ definition of pneumoconiosis.” We agree with Colley & Colley on this issue.
As we stated in Stiltner v. Island Creek Coal Co., 86 F.3d 337, 339 (4th Cir.1996), in the context of a similar regulatory presumption, the rebuttal standard “requires the employer to rule out any causal relationship between the miner’s disability and his coal mine employment by a preponderance of the evidence.” Thus, the ALJ must consider whether the physician found that the miner’s disability was caused by coal dust exposure. Any conclusions by a doctor as to legal versus medical pneumoconiosis, even if in error, were irrelevant to the ALJ’s task, which was to determine whether the evidence was sufficient to overcome the presumption of causation.
Rather than addressing the issue of causation, the ALJ improperly relied on a distinction between medical and legal pneumoconiosis as justification for disregarding certain opinions. The ALJ thus credited the opinions of Drs. Buddington, Garzón, Kanwal, O’Neill, and Schmidt over the conflicting opinions of Drs. Castle, Dahhan, Hansbarger, and Tomashefski. Because the reasons offered by the ALJ for discounting these four opinions go to the issue of establishing pneumoconiosis, not causation, the reasons were insufficient. See Milburn Colliery, 138 F.3d at 528. And as a result, the ALJ failed to consider direct evidence relevant to the issue of causation. See id.; see also 30 U.S.C. § 923(b) (stating that, in the adjudication of claims under the Act, “all relevant evidence shall be considered”). For example, Dr. Hansbarger concluded that exposure to coal dust did not cause Breeding’s emphysema, finding that Breeding’s emphysema was consistent with his 40-year history of cigarette smoking. Dr. Tomashefski concluded that “neither simple eoalworkers’ pneumoconiosis, coal dust exposure nor coal mine employment is a cause of Mr. Breeding’s diffuse emphysema” or his chronic bronchitis and that, instead, these conditions “were caused by his longstanding cigarette smoking habit.” Dr. Dahhan concluded that Breeding’s condition was due to “severe emphysema resulting from his smoking habit” and not to coal mine employment. And Dr. Castle concluded that Breeding “did not have any *568disability during life due to pneumoconiosis, but rather was disabled as a result of tobacco smoke induced pulmonary emphysema.”
Moreover, in crediting other opinions, the ALJ likewise overlooked causation evidence in them, resting instead on the physician’s discussion of the presence of pneumoconiosis or manifestations of it. For example, in crediting Dr. Garzon’s report, the ALJ overlooked the physician’s conclusion that Breeding’s disease was “most likely related to the long history of cigarette smoking.” And in crediting Dr. O’Neill’s report, the ALJ overlooked the physician’s conclusion that Breeding’s “respiratory impairment was caused by his chronic cigarette smoking.” The ALJ failed to consider sufficiently this and other evidence before him on the issue of causation.
Because we are not in a position to reweigh the medical evidence, see Lane, 105 F.3d at 170, we grant the petition on this issue and remand this case to the Board with instructions to remand it to the ALJ for reconsideration of the medical opinion evidence.
PETITION FOR REVIEW DENIED IN PART AND GRANTED IN PART AND CASE REMANDED.