A cohort study of bladder cancer was conducted in a population of California Seventh-day Adventists. Most Seventh-day Adventists use neither tobacco nor alcohol yet experience a large degree of variation in dietary habits. Therefore, diet and other lifestyle habits were evaluated in this unique population. In 1976, 34,198 non-Hispanic white Seventh-day Adventists in California completed a detailed lifestyle questionnaire which included a 51-item food frequency section. This cohort was then followed until the end of 1982 during which time all newly diagnosed malignancies were detected. In order to evaluate the relation between several variables hypothesized to be associated with altered bladder cancer risk, age-, sex-, and smoking-adjusted relative risks (incidence rate ratios) were calculated using the method of Mantel-Haenszel adopted for person-time data. Multivariate analyses were conducted using the Cox Proportional Hazards Regression model. Between the return of the questionnaire and the end of follow-up, there were 52 histologically confirmed bladder cancers detected in the cohort. Increasing age, male gender, and a history of cigarette smoking were all significantly associated with increased bladder cancer risk. In addition, residence in a rural area was associated with significantly increased risk (relative risk (RR) = 1.80) as was high consumption of meat, poultry, and fish (RR = 2.57).