OBJECTIVE To evaluate the influence of a H2 receptor antagonist (cimetidine) on survival in patients with colorectal carcinoma, a randomized, controlled pilot study was performed in three university hospitals in Copenhagen, Denmark. METHODS A total of 192 patients, who had undergone a resection or an exploratory operation for adenocarcinoma of the colon or rectum between May 1988 and May 1991, were enrolled in the study. After a median observation time of 40 months, outcome was noted for each patient concerning cancer-specific mortality rate. RESULTS In patients operated with curative intent (n = 148), no difference was found in cancer-specific mortality between the two treatments. However, a tendency toward reduction in mortality rate was found in patients with curatively operated Dukes Stage C carcinoma (P = 0.11, log-rank test; difference, 29 percent; 90 percent confidence interval, 2 to 57 percent) in the cimetidine-treated group. In patients with disseminated disease no total difference was found between the two treatment groups. CONCLUSIONS Cimetidine does not seem to reduce mortality in patients with colorectal cancer, but there seems to be a tendency toward a survival benefit in patients undergoing surgery for Dukes Stage C carcinoma. Results seem to justify trials in this patient category to reveal a benefit of H2 receptor antagonists in adjuvant therapy of colorectal carcinoma.