Affirmed by published opinion. Chief Judge FABER wrote the opinion, in which Judge KING joined. Judge WIDENER wrote a concurring opinion.
OPINION
FABER, District Judge.Newport News Shipbuilding and Dry Dock Company (“Newport News”) seeks reversal of a decision of the Benefits Review Board (the “BRB”) denying relief under section 8(f) of the Longshore and Harbor Workers’ Compensation Act, 33 U.S.C. §§ 901-950 (“LHWCA”). For the reasons discussed below, we affirm the decision of the BRB.
I.
In April 1997, Herbert E. Winn (“Winn”) was diagnosed with asbestosis. A pulmonary function test revealed a twenty percent permanent partial whole person impairment as measured by the AMA Guides to the Evaluation of Permanent Impairment, 4th ed. His employer, Newport News, admits that he was exposed to asbestos during his employment, but contends that he had a preexisting lung disease which materially and substantially contributed to his ultimate disability.
Winn had a long history of cigarette smoking, admitting to physicians that he smoked a pack of cigarettes a day for nearly forty-five years. As early as 1979, X-rays performed on Winn during his employment with Newport News revealed “increased bronchiovascular markings.” J.A. 4. Winn retired in 1986,* several years before he was diagnosed with asbestosis.
On April 21, 1999, the Administrative Law Judge (“ALJ”) filed his original decision and order denying the employer’s § 8(f) claim. The ALJ found that Newport News had failed to establish the existence of a preexisting permanent partial disability, and had not presented arguments that would support the contribution requirement. More than one year later, on May 9, 2000, the BRB issued a decision and order affirming the ALJ’s findings.
II.
Section 8(f) of the LHWCA, 33 U.S.C. § 908(f), provides in part as follows:
*430In ... cases in which the employee has a permanent partial disability, found not to be due solely to that injury, and such disability is materially and substantially greater than that which would have resulted from the subsequent injury alone, the employer shall provide ... compensation for one hundred and four weeks only.
After the 104-week period has passed, the Office of Workers’ Compensation Programs (“OWCP”), which administers a special fund for that purpose, is responsible for disability payments under § 8(f).
An employer who seeks to limit liability for an employee’s permanent partial disability under § 8(f) must establish three elements: (1) that the ultimate disability is caused in part by a preexisting partial disability; (2) that the preexisting disability was manifest to the employer prior to the work-related injury; and (3) that the ultimate disability materially and substantially exceeded the disability that would have resulted from the work-related injury alone, in the absence of the preexisting condition. See Director, OWCP v. Newport News Shipbuilding and Dry Dock Co. (Carmines), 138 F.3d 134, 138-39 (4th Cir.1998). See also Director, OWCP v. Newport News Shipbuilding and Dry Dock Co. (Harcum), 8 F.3d 175, 182-83 (4th Cir.1993), aff'd on other grounds, 514 U.S. 122, 115 S.Ct. 1278, 131 L.Ed.2d 160 (1995).
The LHWCA sets out the applicable standard of review for this case in § 921(b)(3). When under review by the BRB, the ALJ’s factual findings are deemed conclusive “if supported by substantial evidence in the record considered as a whole.” 33 U.S.C. § 921(b)(3). This court has held that substantial evidence is “more than a scintilla but less than a preponderance.” Elliott v. Adm’r, Animal & Plant Health Inspection Serv., 990 F.2d 140, 144 (4th Cir.1993).
Additionally, as stated in Newport News Shipbuilding & Dry Dock Co. v. Tann, 841 F.2d 540, 543 (4th Cir.1988), the ALJ’s findings “may not be disregarded on the basis that other inferences might have been more reasonable.” Rather, “deference must be given the fact-finder’s inferences and credibility assessments, and we have emphasized the scope of review of ALJ findings is limited.” Id. Legal determinations are reviewed de novo. This includes the standard by which the ALJ determines entitlement to § 8(f) relief. See Carmines, 138 F.3d at 141. We review the Board’s decision “for errors of law and to ascertain whether the Board adhered to its statutorily-mandated standard for reviewing the ALJ’s factual findings.” Newport News Shipbuilding & Dry Dock Co. v. Tann, 841 F.2d 540, 543 (4th Cir.1988).
III.
In the instant case, the first and third elements under § 8(f), preexisting disability and contribution, respectively, are in question. The second element, manifestation, is not required in cases where the worker suffers from a post-retirement occupational disease, as is the case with Winn. The ALJ discussed the evidence and found that Newport News had not established that Winn had a preexisting disability. J.A. at 37. However, he also found that even had a preexisting disability been proven, Newport News had not presented evidence sufficient to meet the contribution criteria set forth by this court in Carmines. Id. The BRB agreed, although focusing its opinion solely on Newport News’ failure to establish contribution.
To meet the contribution element, an employer must quantify the type and *431extent of the disability the claimant would have suffered without the preexisting condition. See Carmines, 138 F.3d 134, 139 (citing Harcum). Such a showing requires quantification of the level of impairment that would ensue from the work-related injury alone. “In other words, an employer must present evidence of the type and extent of disability that the claimant would suffer if not previously disabled when injured by the same work-related injury.” Id.
To satisfy the contribution element in this case, Newport News relied upon the parties’ stipulation that Winn has a twenty percent permanent partial disability, and on the opinions of three doctors, including Dr. James Reid, a company doctor, and Drs. Charles Donlan and Ricardo Guardia, neither of whom treated Winn for his condition. Reid stated that if Winn had only the work-related asbestosis, his impairment rating would be at least ten percent less. Dr. Reid explained his conclusion and the basis for it as follows in September 1997:
1. The Shipyard clinic read Mr. Winn’s chest x-rays on April 9, 1979 and again on April 27, 1982 to show “Inc BVM”— increase bronchovascular markings (Exhibit 1). These are the x-ray findings indicative of chronic obstructive pulmonary disease (“COPD”). The Shipyard clinic interpreted Mr. Winn’s pulmonary function tests in 1981, 1982, 1983, and 1985 to show “mild SAO” — small airways obstruction (Exhibits 2 and 3). In a long time cigarette smoker such as Mr. Winn, these chest x-ray findings and pulmonary function tests were diagnostic for COPD. Thus, by 1985, Mr. Winn was known to have COPD with approximately 10% impairment.
2. Thus, if Mr. Winn did not have COPD, and only his alleged asbestosis, his AMA rating would be at least 10% less.
J.A. at 9.
In December 1998, Dr. Donlan wrote to counsel for Newport News and stated as follows:
Thank you for your recent letter regarding Mr. Wynne [sic]. I have reviewed Dr. Reed’s [sic] report as well as my office records. I have not treated Mr. Wynne [sic] but did see him for a pulmonary evaluation in September 1997. I concluded that he did have pulmonary asbestosis. He did have a history of occupational asbestosis exposure with minimal x-ray abnormalities, presence of bilateral crackles, and reduction in diffusion capacity.
He did have a cough and mild shortness of breath on exertional activities. I felt that this was likely related to chronic bronchitis. His pulmonary function tests performed at DePaul Hospital September 23, 1997 did show mild obstructive impairment. Diffusion capacity was mildly reduced to 75% of predicted.
Using AMA guidelines, I would place his impairment as Class II, 10%. I think the majority of this impairment would.be secondary to chronic bronchitis from cigarette smoking. I would agree with Dr. Reed [sic] that the preexisting disease of chronic bronchitis contributes to his overall impairment. I would agree with Dr. Reed [sic] that had he not had chronic bronchitis that his impairment would be less.
J.A. at 10.
In January of 1999, Dr. Donlan provided another letter in response to an inquiry by counsel for Newport News. In this letter, Dr. Donlan writes, “I will attempt to answer your question in regards to Mr. Wynn’s [sic] overall impairment. I would conclude that his overall impairment would be 4% had he had asbestosis alone.” J.A. *432at 11. Dr. Guardia, who saw Winn in 1988 following an auto accident, but never treated Winn for any other condition, reviewed Winn’s record at the clinic, as well as the reports of Drs. Donlan and Reid, and merely agreed that “had Mr. Wynn [sic] not been a smoker, his disability would have been much less.” J.A. at 12.
The BRB concluded that the evidence of contribution fell short of the standard for § 8(f) relief established by this court in Carmines. Carmines was exposed to asbestos at the Newport News facility over a thirty-year period. See Carmines, 138 F.3d at 137. He was diagnosed with pulmonary asbestosis in 1990 and deemed to have a twenty-five percent to thirty percent impairment of the whole person. See id. He also had substantial preexisting scarring of the lungs due to pleurisy. See id. at 139. Dr. Hall, a company physician for Newport News, opined on the basis of old x-rays that Carmines’ preexisting disability accounted for about eighteen percentage points of the total disability, which was twenty-eight percent. See id. at 140. He concluded that Carmines’ asbestosis accounted for the difference — ten percent. See id.
This court determined in Carmines that Dr. Hall’s evidence was insufficient to meet the contribution requirement and show the claimant’s total disability to be materially and substantially greater than the disability resulting from the work-related condition alone. See Carmines, 138 F.3d at 134, 142 (4th Cir.1998). The court pointed out that Carmines’ asbestosis standing alone could account for all of his disability. See id. at 143. In the absence of expert evaluation of the degree of disability covered by the asbestosis separate and apart from the impact of the preexisting condition, the court said, there was no basis for a comparison which would reveal whether the prior injury materially and substantially contributed to the total impairment. See id. at 143^14. To satisfy the third requirement under § 8(f), the court held that, “an employer must quantify the type and extent of the disability that the claimant would have suffered without the preexisting condition.” Id. at 139. It is not enough, as Dr. Hall did in Carmines, to simply calculate the.total current disability and subtract from it the disability resulting from the preexisting condition.
As we have noted in other § 8(f) cases, competing policy goals are always at stake in these cases. The purpose of § 8(f) is to encourage reemployment of disabled workers and to prevent an employer from being additionally hable when a previously disabled employee is injured. See Lawson v. Suwanee Fruit & S.S. Co., 336 U.S. 198, 201, 69 S.Ct. 503, 93 L.Ed. 611 (1949); Carmines, 138 F.3d at 139. However, § 8(f) should not produce a windfall to employers who should be held responsible for the work-related injuries and illnesses which befall previously disabled employees. See Carmines, 138 F.3d at 139. Additional difficulties result from the frequent appearances of company doctors providing opinions for an employer who has an incentive to try to get as many cases as possible dumped into the special fund. In this context, the rigorous standard of Carmines makes good sense and is supported by cases from other circuits. In Sealand Terminals, Inc. v. Gasparic, 7 F.3d 321, 323 (2d Cir.1993) (per curiam), the court held that § 8(f) relief is precluded unless the evidence establishes that the work-related injury alone would not have caused the workers’ total disability. The United States Court of Appeals for the First Circuit, in Director, OWCP v. Bath Iron Works Corp., 129 F.3d 45 (1st Cir.1997), followed this court’s opinion in Harcum, and held that, to justify § 8(f) relief, an employer must show the degree of dis*433ability attributable to the work-related injury alone, “so that this amount may be compared to the total percentage of the partial disability for which coverage under the LHWCA is sought.” 129 F.3d at 51; see also Two “R” Drilling Co., Inc. v. Director, OWCP, 894 F.2d 748 (5th Cir.1990).
Applying the Carmines standard here, we find the evidence insufficient to support § 8(f) relief. We agree with the BRB that Dr. Reid’s methodology, as well as that of Dr. Guardia, fails to satisfy Carmines. Dr. Reid concluded that Winn’s asbestosis, without the preexisting condition, would result in an AMA rating of at least ten percent less than the rating which results from the two conditions combined. He thus places the degree of disability from the asbestosis alone at less than the total disability, but gets there by using a process much like the “subtraction” process disfavored by this court in Carmines. As the BRB found, it appears that Dr. Reid merely subtracted the extent of disability resulting from the preexisting disability from the extent of the current disability. This method is legally insufficient under Carmines to establish that a claimant’s preexisting disability is materially and substantially greater than the disability due to the final injury alone.
Similarly, Dr. Guardia’s opinion, which merely states that if Winn had not been a smoker, his disability would have been “much less” is also legally insufficient. This opinion does not attempt to quantify the level of impairment that would result from the work related injury alone, as is required by Harcum. Dr. Guardia’s short opinion is similar to that of a doctor who reviews a claimant’s medical records and then uses a “check-off’ form to indicate his agreement with the assessment of another doctor. These forms are used in many cases, and are usually given little weight by ALJs. Further, Dr. Guardia never treated Winn, and only examined him once several years earlierafter an auto accident. The BRB and the ALJ were correct in giving little weight to this opinion.
An ALJ may not “merely credulously accept the assertions of the parties or their representatives, but must examine the logic of their conclusions and evaluate the evidence upon which their conclusions are based.” Carmines, 138 F.3d at 140. The ALJ in the instant case performed this duty when he found that the opinions of Drs. Donlan and Guardia had a questionable basis. The ALJ correctly noted that Dr. Guardia’s report did not mention which reports were reviewed and did not provide a basis for the conclusion reached. J.A. at 37. In his opinion, the ALJ set forth the full text of the doctors’ opinions, in which Dr. Donlan and Dr. Guardia revealed that they had not treated Winn, and had little contact with him. Id. The ALJ further notes the discrepancy between the numbers reached by Dr. Reid and Dr. Donlan, and that testing results from Dr. Donlan were not submitted. Id. For all of these reasons, we feel that the BRB was correct in upholding the decision of the ALJ, and finding that the employer failed to establish the contribution element. Based upon this ruling, the issue of preexisting permanent partial disability need not be addressed.
Accordingly, the decision of the BRB is
AFFIRMED.
Although the year varies in several documents found in the Joint Appendix, the parties previously stipulated that it was 1986.