A case-control study was conducted to investigate occupational risks for colon cancer. The Missouri Cancer Registry was used to identify a total of 1,993 white male cases and 9,965 age-matched cancer controls diagnosed between 1984 and 1987. Effect estimates were calculated for specific occupations and industries and for three levels of occupational physical activity. Odds ratios (ORs) and 95% confidence intervals (Cls) were also calculated according to anatomic subsite of colon cancer. Elevated risks were identified for workers in several occupations and industries including printing machine operators (OR = 1.9, 95% Cl 1.0-3.3), workers in food manufacturing (OR = 1.6, 95% Cl 1.2-2.2), communications workers (OR = 1.8, 95% Cl 1.0-3.3), and workers employed in the petroleum product trade (OR = 2.7, 95% Cl 1.0-7.2). Previous findings of excess risk among selected occupations of higher socioeconomic status were confirmed. Analyses by anatomic subsite showed that excess risk for an occupational group was usually confined to a single subsite, which suggests some specificity of effect. For example, elevated risk among males employed in manufacturing-related industries was shown primarily for cancer of the cecum. The findings by level of occupational physical activity corroborated recent reports of increased colon cancer risk associated with employment in sedentary occupations. Excess risk (OR = 1.4, 95% Cl 1.0-1.9) was shown among males employed in sedentary jobs and an inverse linear trend in risk was shown according to level of occupational physical activity. Risk associated with sedentary occupations was highest for cancer of the cecum (OR = 2.1, 95% Cl 1.1-4.0). Despite the limitations of the cancer registry-collected data, some associations were observed that deserve further study.