To assess the relationship of smoking and coffee, tea, and alcohol intake to the risk of cancer of the exocrine pancreas, analyses were performed using data from a prospective cohort study of 33,976 postmenopausal Iowa women who responded to a mailed questionnaire in 1986 and were followed through 1994 for cancer incidence and total mortality. At baseline, information on cigarette smoking, consumption of tea, coffee, and alcoholic beverages, and other dietary and lifestyle factors was obtained. Age-adjusted relative risks of pancreatic cancer (n = 66 cases) showed a dose-response association with smoking. Those with fewer than 20 pack-years and those with 20 or more pack-years of smoking exposure were 1.14 (95% confidence interval, 0.53-2.45) and 1.92 (95% confidence interval, 1.12-2.30) times more likely, respectively, to develop pancreatic cancer than were nonsmokers. Current smokers were twice as likely as were nonsmokers to develop pancreatic cancer. Relative risks of pancreatic cancer increased with the amount of alcohol consumed (Ptrend = 0.11) after adjustment for age, smoking status, and pack-years of smoking. Relative risks of pancreatic cancer according to alcoholic beverage intake were as strong among never-smokers as they were in the total cohort. After the data were adjusted for age, smoking status, and pack-years of smoking, there was a statistically significant 2-fold (95% confidence interval, 1.08-4.30) elevated risk of pancreatic cancer for those who drank > 17.5 cups of coffee per week, compared to those who consumed < 7 cups/week; among never-smokers, the relative risks across coffee intake categories were still positive but were attenuated somewhat (P trend = 0.17). Tea intake was not related to cancer incidence. In summary, these findings provide evidence of an association of both alcoholic beverage and coffee consumption with pancreatic cancer incidence that is independent of age and cigarette smoking.