Epidemiologic data suggest that consumption of whole-grain products may be inversely associated with risk of pancreatic cancer. Grain intake was examined in a population-based case-control study of pancreatic cancer in the San Francisco Bay Area (1995-1999). A 131-item semiquantitative food frequency questionnaire was administered to 532 cases and 1,701 controls. Odds ratios and 95% confidence intervals were computed as estimates of relative risk. Persons who consumed > or =2 servings of whole grains daily had a lower risk of pancreatic cancer than persons who consumed <1 serving/day (odds ratio (OR) = 0.60, 95% confidence interval (CI): 0.31, 1.2; trend-p = 0.04). Similar results were observed for brown rice (OR = 0.72, 95% CI: 0.44, 1.2; trend-p = 0.01) and tortillas (OR = 0.56, 95% CI: 0.35, 0.89; trend-p = 0.02). Consumption of doughnuts (> or =2 servings/week vs. <1 serving/month) conferred increased risk (OR = 1.8, 95% CI: 1.2, 2.7; trend-p = 0.003). Consumption of cooked breakfast cereals (> or =2 servings/week vs. <1 serving/month) was positively associated with risk (for oatmeal/oat bran, OR = 1.3, 95% CI: 1.0, 1.7; for other cooked breakfast cereals, OR = 2.1, 95% CI: 1.4, 3.3). Dietary fiber was inversely associated with risk (for highest quartile vs. lowest, OR = 0.65, 95% CI: 0.47, 0.89; trend-p = 0.02). These data provide some support for the hypothesis that consuming more whole-grain or high-fiber foods may reduce the risk of pancreatic cancer. Refined grains were not associated with risk.