Regular continuing nonsteroidal anti-inflammatory drug (NSAID) use has been associated with a reduction in risk of large bowel cancer in many studies, including our Case-Control Surveillance Study of medication use and cancer risk. We assessed the relation of NSAID use to the risk of digestive cancers at sites other than the large bowel in this database. Nurse-interviewers administered questionnaires to patients admitted to hospitals in four centers from 1977 to 1998. Cases comprised 1149 patients with cancers of the pancreas (n = 504), stomach (n = 254), esophagus (n = 215), gallbladder (n = 125), or liver (n = 51). Controls were 5952 patients admitted for trauma or acute infection. History of NSAID use was elicited by questions about indications for use. Multiple logistic regression models were used to calculate odds ratios (ORs) for categories of regular NSAID use (at least 4 days/week for at least 3 months) relative to never use. The OR for regular use initiated at least 1 year before admission and continuing into that year was reduced for stomach cancer (OR = 0.3; 95% confidence interval, 0.1-0.6) and was compatible with 1.0 for other sites. The ORs for regular continuing use of at least 5 years duration were < 1.0 for cancers of the stomach, pancreas, esophagus, and gallbladder but were statistically significant only for stomach cancer. These data suggest that regular continuing NSAID use may be associated with reduced risk of stomach cancer. For the other sites, the data are consistent with no effect of NSAID use.