The effectiveness and selectivity of proglumide, a putative cholecystokinin/gastrin receptor antagonist in vitro, were examined on gastric acid and pancreatic secretion in vivo. Gastric secretion was measured in conscious dogs in the basal state and during infusion of pentagastrin, histamine, or bethanechol, alone or in combination with proglumide (300 mg/kg . h). Pancreatic secretion was measured in anesthetized rats in response to cholecystokinin-octapeptide or secretin, alone or in combination with proglumide (100 mg/kg). Proglumide inhibited pentagastrin-stimulated secretion but had no effect on basal, histamine-stimulated, or bethanechol-stimulated gastric acid secretion. Inhibition of pentagastrin-stimulated secretion was of the competitive type. An apparent inhibitory constant was calculated to be 300 mg/kg . h; this dose is capable of eliciting plasma concentrations of approximately 1 mM. This estimate corresponds closely to that derived from measurements in isolated canine parietal cells. Proglumide also inhibited cholecystokinin-stimulated but not secretin-stimulated pancreatic secretion. The lack of effect of proglumide on basal, histamine-stimulated, or bethanechol-stimulated gastric acid secretion implies that background gastrin has no direct or synergistic influence on stimulation by other secretagogues. The selective effect of gastrin receptor antagonists contrasts with the effectiveness of muscarinic and histamine H2-receptor antagonists against secretion induced by all types of stimulants. Accordingly, the antisecretory potential of gastrin receptor antagonists is confined to digestive secretion when the effect of gastrin is optimal. Their potential as antitrophic agents in duodenal ulcer disease, however, has not been explored yet.