Gastric cancer remains a disease with a very poor prognosis, and there is no safe and effective form of therapy for advanced disease. Evidence is now abundant to show that gastrin stimulates the growth of both gastric and colorectal cancer cells in vitro and in vivo, and that blockade of gastrin receptors can prolong survival in xenografted nude mice. We have thus performed a randomized, controlled study of the gastrin/cholecystokinin receptor antagonist proglumide as therapy in 110 patients with gastric carcinoma. Proglumide had no overall effect on survival (Mantel-Cox statistic = 0.5, P = 0.48). The 95% confidence interval for the proglumide treated group was 260 to 474 days compared to 230 to 372 days for the control group. No significant difference was seen with proglumide, which has a relatively low affinity with the gastrin receptor and also has partial agonist activity. Drugs that are far more specific and potent gastrin receptor antagonists are becoming available, which may have a greater effect on survival, and further clinical trials of such compounds are clearly indicated to determine the efficacy of hormonal control of gastrointestinal malignancy.