Although early studies attributed an important role to acid in the pathogenesis of duodenal ulcer pain, recent reports are conflicting. The aim of the present study is to determine whether direct acidification of the duodenal ulcer crater in symptomatic patients reproduces ucler pain. Patients with endoscopically diagnosed duodenal ulcers were studied. No premedication or sedation was given. A washing tube was passed via the endoscope and 0.1 N hydrochloric acid as well as normal saline were sequentially administered on to the ulcer crater, the sequence of infusion being randomised and double blind. Forty patients were studied. Sixteen developed typical ulcer pain during acid infusion compared with four with saline (p less than 0.005). Ten patients who developed pain on acid were rechallenged with acid after their pains disappeared. Typical pain recurred in all. Twenty patients without duodenal ulcer did not develop pain when 200 ml of 0.1 N hydrochloric acid was infused into the duodenum. Acid therefore appears to have a definite role in the pathogenesis of duodenal ulcer pain although other factors may also be important.