Four dogs with chronic gastric fistulas were give intravenous bombesin nonapeptide (B9), ranatensin, and litorin by constant infusion for 90 min at 1.2 micrograms x kg-1 on separate days. A dose response study with substance P (1.5, 3.0, 60, 18 and 54 micrograms x kg-1 x h-1) was also carried out and all tests compared to a standard protein meal (10g x kg-1). Plasma gastrin and PP were measured by radioimmunoassay and gastric acid by autobiuret titration. Substance P failed to stimulate gastric acid secretion or release either pancreatic polypeptide (PP) or gastrin. Basal gastrin levels were 8 +/-2 fmol/ml. The peak increment of gastrin released by bombesin was 95 +/- 16, ranatensin 22 +/- 6, litorin 18 +/- 4, and meal 39 +/- 5 fmol/ml. Bombesin caused significantly greater release of gastrin than a meal, litorin or ranatensin (P less than 0.01). Basal gastric secretion was 23 +/- 4 microequiv./min. B9 produced a peak acid secretion of 356 +/- 124 muequiv./min. There was no significant difference between the bombesin-like peptides (P less than 0.01). Basal plasma PP was 38 +/- 12 fmol/ml. B9 produced a peak PP increment of 600 +/- 50, litorin 137 +/- 36, ranatensin 98 +/- 11, and a meal 305 +/- 58 fmol/ml. B9 released significantly more PP than either litorin of ranatensin (P less than 0.01). The different amino acid sequences of the peptides are probably responsible for their potency. The substitution of a penultimate phenylalanine residue in litorin and ranatensin for leucine in bombesin does not prevent PP or gastrin release by bombesin-like peptides. Since bombesin-like peptides are widely distributed in the gastrointestinal tract of man and stimulate both acid and gut hormone secretion, it is possible that they might play a physiological role in the modulation of gastrointestinal function.