International and interethnic differences in prostate cancer incidence suggest an environmental aetiology, and lifestyle and socio-economic factors have been studied, but with divergent results. Information on a cohort of 22,895 Norwegian men aged 40 years and more was obtained from a health examination and two self-administered questionnaires. Information on incident cases of prostate cancer was made available from the Cancer Registry. We used the Cox proportional hazards model to calculate incidence rate ratios as estimates of the relative risk (RR) with 95% confidence interval (CI). Reported P-values are two-sided. During a mean follow-up of 9.3 years, 644 cases were diagnosed. Risk was elevated among men in occupations of high compared to low socio-economic status (RR = 1.30; 95% CI 1.05-1.61), and among men with high education compared to the least educated (RR = 1.56; 95% CI 1.11-2.19). A RR of 1.56 (95% CI 0.97-2.44) suggests a higher risk among divorced or separated men, compared with married men. We also found indications of a weak negative association with leisure-time physical activity (RR = 0.80; 95% CI 0.62-1.03 for high vs low activity), a weak positive association with increasing number of cigarettes (P = 0.046), while alcohol consumption was not related to the risk of prostate cancer. These results show that high socio-economic status is associated with increased risk of prostate cancer, and that divorced or separated men might be at higher risk than married men. Data from this study also indicate that high levels of physical activity may reduce prostate cancer risk.