In this Social Security disability case, Donald Smallwood maintains that the Commissioner improperly disregarded the opinion of a treating physician. We affirm.

* Smallwood seeks disability benefits starting September 14, 1991, the date of his second heart attack. Smallwood, who quit his last job two weeks before his second heart attack, has not worked since. Smallwood's heart has been damaged by coronary disease and heart attacks, but not sufficiently to meet the standard for automatic disability.

In addition to his cardiac problems, Smallwood is a smoker (1 1/2 packs per day at the time of the hearing) and has a lifelong history of alcohol abuse, including several failed attempts at treatment. The administrative law judge (ALJ) found that Smallwood's drinking was to at least some extent controllable and thus not alone disabling, a finding that we conclude is supported by substantial evidence.

It is undisputed that Smallwood cannot return to his past relevant work. This case turns on whether the Commissioner has shown that Smallwood has the residual functional capacity to work at other jobs. The parties offer differing interpretations of one key residual functional capacity assessment performed by treating cardiologist Dr. David Lemon. The ALJ for the most part accepted Dr. Lemon's opinions on Smallwood's residual functional capacity. However, the ALJ found Dr. Lemon's opinion that Smallwood could work only four hours per day to be inconsistent with the rest of Dr. Lemon's report. (It is undisputed that Smallwood would be disabled if he were limited to four-hour days.)

The ALJ rejected Dr. Lemon's four-hour day conclusion because Dr. Lemon also opined that Smallwood could stand and walk for one hour without a break and could sit indefinitely without a break. Also important to the ALJ's conclusion that Smallwood could work more than four hours a day was the ALJ's finding that Dr. Lemon had not corroborated Smallwood's claim to need a nap in the afternoon. Id. In approving the ALJ's decision, the district court added a related reason for rejecting Dr. Lemon's four-hour conclusion: that an opinion regarding time limits on a claimant's work is "vocational--not medical." The government strongly presses this position before us.

The assumption that physicians cannot opine as to the hours a claimant can work is wrong. Physicians regularly make such assessments. E.g., Walston v. Sullivan, 956 F.2d 768, 770 (8th Cir.1992); Rodriguez v. Bowen, 876 F.2d 759, 760-61, 763 (9th Cir.1989); Zenker v. Bowen, 872 F.2d 268, 271-72 (8th Cir.1989). In fact, medical opinions on how much work a claimant can do are not only allowed, but encouraged. 20 C.F.R. Sec. 404.1513(b) ("Medical reports should include ... [a] statement about what you [the claimant] can still do despite your impairment(s)....") (emphasis added).

The government cites Nelson v. Sullivan, 946 F.2d 1314, 1316-17 (8th Cir.1991) (per curiam) and Turley v. Sullivan, 939 F.2d 524, 527 (8th Cir.1991) for the proposition that physicians normally cannot give vocational testimony. But neither case is apropos here because Dr. Lemon did not address vocational matters. The treating physicians in both Nelson and Turley gave opinions as to whether the claimant could find work or be gainfully employed. Nelson, 946 F.2d at 1316; Turley, 939 F.2d at 526. This type of conclusion is outside the medical province: it is for a vocational expert to take into account medical limitations, including opinions as to work time limits, and offer an opinion on the ultimate question whether a claimant is capable of gainful employment. In this case, the only ultimate opinions in the record as to whether Smallwood could be gainfully employed are the vocational expert's.

Since Dr. Lemon did not overstep his medical role in stating his opinion that Smallwood was limited to four-hour days, and since Dr. Lemon is Smallwood's treating cardiologist, his opinion on the matter controls if it "is well-supported by medically acceptable ... diagnostic techniques and is not inconsistent with the other substantial evidence" in the record. 20 C.F.R. Sec. 404.1527(d)(2). As there is no contention that Dr. Lemon's diagnostic techniques were not acceptable, this case resolves to whether Dr. Lemon's opinion is either not well-supported or is contradicted by substantial evidence.

Dr. Lemon's own residual functional capacity questionnaire is itself inconsistent with his four-hour day conclusion: when asked how many pounds Smallwood could carry "on a practical day-to-day, sustained basis," Dr. Lemon simply wrote that Smallwood could frequently handle 10 pounds, with no mention of any time limit. When asked if Smallwood would need to lie down during an eight-hour workday, Dr. Lemon checked the "No" box. Dr. Lemon's failure to add qualifications to these answers invites the implication that Smallwood could work full-time. This implication is not dispelled when Dr. Lemon turns to the final question of how many hours Smallwood can work each day. Dr. Lemon simply filled in the "hrs./day" box with a "4" with no explanatory comment. Dr. Lemon's bare assertion that Smallwood could work only four hours is inconsistent with the rest of the questionnaire, and this important inconsistency is nowhere explained by Dr. Lemon.

Smallwood argues that Dr. Lemon's four-hour workday opinion is supported by Dr. Lemon's reference in the questionnaire to Smallwood's two heart attacks, coronary disease, and limited cardiac reserve. But Dr. Lemon's reference to these medical conditions was made in response to a different question and is nowhere tied by Dr. Lemon to his four-hour day opinion. It is simply not clear what factors Dr. Lemon relied on in deriving his four-hour workday opinion. Dr. Lemon's opinion is the kind of conclusory statement that cannot be accorded the deference to which well-supported treating physician determinations are entitled. Frankl v. Shalala, 47 F.3d 935, 938 n. 1 (8th Cir.1995); 20 C.F.R. Sec. 404.1527(d)(2). When faced with a conclusory opinion by a treating physician, the Commissioner need only come forth with "some medical evidence" that the claimant can work. Frankl, 47 F.3d at 938. Residual functional capacity assessments by non-treating physicians can constitute the requisite substantial evidence. Cf. id. (RFC assessments did not constitute substantial evidence only because done before claimant's condition deteriorated).

Like Dr. Lemon's report, two other residual functional capacity assessments by physicians imply that Smallwood can work more than four hours per day. These assessments stated that Smallwood could with normal breaks either stand, walk or sit for about six hours in each eight-hour workday. This implies that Smallwood could work six hours a day, if not more. ("[A]bout 6 hours" is the longest time choice offered on the questionnaires these doctors used. Unlike Dr. Lemon's questionnaire, the one used by the other doctors did not directly ask how many hours per day Smallwood could work.)

Dr. Lemon's opinion on the issue of how many hours Smallwood can work is inconsistent with other aspects of his own assessment of Smallwood's residual capacity; taken as a whole, Dr. Lemon's residual capacity report itself implies that Smallwood can work more than part-time. Also, Dr. Lemon's assessment of residual capacity jibes almost perfectly with those of two other doctors except on the isolated issue of hours per day, and these other residual functional capacity assessments clearly contemplate that Smallwood can work a minimum of six hours per day. For these reasons, we find that the treating physician's conclusory opinion as to the necessity of a four-hour workday is inconsistent with other substantial evidence in the record. Accordingly, we affirm the judgment denying the claim for benefits.