Recently smoking has been increasingly implicated as a possible risk factor for low-back pain. One explanation for this finding is confounding by occupation. To investigate this possibility, the relationship between smoking and self-reported back pain was studied within 13 occupations. A relationship between smoking and back pain was observed only in occupations that require physical exertion. The relationship between smoking and other musculoskeletal pain also was explored. Pain in the extremities turned out to be related more clearly to smoking than to pain in the neck or the back. This suggests confounding or a general influence of smoking on pain. It is concluded that prevention of back pain could be a beneficial side-effect of anti-smoking campaigns. However, the prime target for prevention of low-back pain would have to be other factors.