From 1984 to 1988 a population-based case-control study was carried out in the Netherlands, in collaboration with the International Agency for Research on Cancer, to examine the possible relationship between the habitual lifetime consumption of alcohol, coffee and tea and exocrine pancreatic carcinoma in 176 cases and 487 controls. An interviewer-administered questionnaire was used to ascertain major life events and obtain estimates of consumption (ever-never) and frequency of consumption throughout life. Logistic regression analyses yielded odds ratios adjusted for age, sex, response status, smoking, dietary intake of energy and vegetables and of alcoholic or non-alcoholic drinks. When compared with data from non-drinkers, the cumulative lifetime consumption of all types of alcohol in grams of ethanol (ORs 1.00, 0.97, 0.93, 1.25, p trend 0.55), beer, spirits, red wine and fortified wine was not related to risk. The consumption of white wine was inversely associated with risk (OR 0.41, 95% CI 0.24-0.70). The uniformly reduced risk estimates for the lifetime number of drinks of white wine were based on small numbers (ORs 1.00, 0.44, 0.25, 0.40, p trend 0.001). When compared with data from non-drinkers, our findings suggest an inverse dose-response relationship for the lifetime consumption of coffee (ORs 1.00, 0.72, 0.37, 0.58, p trend 0.06), whereas lifetime consumption of tea and of ground, instant and decaffeinated coffee was not associated with risk. The absence of an effect of lifetime consumption of decaffeinated coffee may be due to the small numbers of subjects. These results further strengthen existing evidence against a positive association between consumption as well as lifetime consumption of (sources of) alcohol, tea or coffee and the development of exocrine pancreatic cancer.