This study analyzes the relationships of cigarette smoking and smoking histories to employment status and occupation. Data from a large, representative sample of U.S. adults in 1985 were analyzed separately for white men, white women, black men, and black women, with controls for age, education, and marital status included in all analyses. It has been hypothesized that women who adopt traditional male roles are more likely to become smokers. However, our data indicate that participation of women in the labor force or employment of women in male-dominated occupations have had little or no effect on women's smoking adoption or cessation. Unemployed men and women (those who were not employed but were seeking work) were substantially more likely than employed men and women to be smokers. Among employed whites, the percentage of smokers was high for industrial manual workers and service workers and low for professionals and farmers. Similar differences were observed in the proportion who had ever adopted smoking. These differences in smoking adoption had begun at young ages, before most people have adopted adult roles, which suggests that the differences in smoking adoption were not caused by unemployment or occupation. Rather, it appears that certain personal characteristics or early experiences influenced both smoking adoption and adult unemployment or occupation. In addition, unemployment may decrease smoking cessation, and employment in professional occupations may increase smoking cessation.