A study was conducted to determine the associations between occupational status and detailed measures of smoking exposure: ever vs. never smoking, type of tobacco used, current vs. ex-cigarette smoking, amount smoked, age began, and tar yield of the usual brand smoked. Date were obtained between the years 1977 and 1979 as part of a large-scale epidemiological study of tobacco use. Subjects interviewed were 2,528 white males aged 41 to 70 while they were patients in hospitals located in five U.S. cities. Thirty-eight percent of the sample had cancer of a site not previously linked with use of tobacco and 62% had non-cancer conditions also unrelated to tobacco exposure. It was found that men in professional and technical occupations showed a markedly higher rate of never smoking than did men from all other occupations who showed only slight differences among themselves. The intensity of other cigarette smoking variables--current vs. ex-smoking, age began, and tar yield of cigarette smoked (but not number per day)--varied significantly by occupational level, with higher levels of smoking intensity observed among men in blue-collar than among those in white-collar occupations. These findings indicate that an appropriate evaluation of an occupationally related disease also affected by smoking must include detailed and comprehensive smoking data. Moreover, it can be expected that men in occupations associated with higher indices of smoking intensity will have higher rates of tobacco-related diseases than those in occupations with lower cigarette intensity exposures.