The effect of cimetidine (1 g daily) and placebo was studied in a controlled clinical trial comprising 50 patients with non-ulcer dyspepsia in whom an organic abnormality responsible for the dyspeptic symptoms was not disclosed by a standardized and extensive examination programme. Reduction of symptoms occurred in 13 (54%) out of 24 patients treated with cimetidine and in 16 (62%) out of 26 treated with placebo. The difference was insignificant, as were the alterations in the individual dyspeptic symptoms between the groups. Only 6 patients (25%) on cimetidine and 8 (31%) on placebo treatment had a total relief of symptoms. Of these, all cimetidine-treated patients remained free from symptoms during the successive 3-month observation period, while the dyspeptic symptoms relapsed in 3 (38%) placebo-treated patients. Subsequent resumption of placebo treatment reduced the symptoms in all 3 patients, but only one became free from symptoms. Cimetidine does not seem to be superior to placebo in the treatment of non-ulcer dyspepsia in patients without any previous history of ulcer or without any sign on endoscopy of an active or previous ulcer disease.