Although the etiologic relation between nonsteroidal antiinflammatory drug (NSAID) use and gastrointestinal lesions is well documented, newly introduced NSAIDs deserve a fresh examination for their risk/benefit ratio. To estimate the association between consumption of ketorolac and the occurrence of gastroduodenal lesions, we conducted a case-control study. The study population comprised 600 outpatients with a confirmed endoscopic diagnosis of ulcer and erosion in 1991 and 1992 and 6,000 community controls matched by age and sex. We retrieved the prescription history through a computerized prescription monitoring system. We defined exposure to each study drug as “current” (month of endoscopy and preceding month), “recent” (second or third month preceding endoscopy), and “past” (fourth to sixth month preceding endoscopy). Current users of NSAIDs showed a 30% increase in the incidence of gastroduodenal lesions [odds ratio (OR) = 1.3; 95% confidence interval (CI) = 0.98–1.8] after adjustment for recent or past use of any NSAID, recent or past gastrotoxic therapy, recent or past use of gastroprotective drugs, and recent or past use of any other drug. Among NSAIDs, ketorolac was the only one showing a distinctly elevated risk of gastroduodenal lesions (OR = 4.2; 95% CI = 1.9–9.4). Current use of any NSAID was associated with almost a doubling of risk for ulcer alone (OR = 1.9; 95% CI = 1.3–3.0); no elevation in risk was found for erosions. The adjusted relative risk for ulcer associated with current use of ketorolac was 9.8 (95% CI = 3.4–28.1). Recent and past use of NSAIDs does not increase the risk of ulcer. The use of ketorolac appears to carry a greater gastrotoxicity than other NSAIDs.