1. The effect of bombesin, a neurogastrointestinal peptide, on basal and stimulated insulin release was studied in man. 2. Two different stimuli were used, hyperglycaemic (20 g glucose) and hypoglycaemic (1 g tolbutamide). They were injected intravenously to two groups of male healthy volunteers during saline or bombesin (5 ng kg-1 min-1 for 60 min) infusion. 3. The peptide had no significant effect on basal levels of glucose and insulin. However, the insulin response to intravenous glucose was strongly potentiated by bombesin, the integrated insulin response being 2.23 +/- 0.59 mu ml-1 . 90 min and 0.98 +/- 0.19 mu ml-1 . 90 min during infusion of bombesin and saline, respectively (P less than 0.05). The behaviour of plasma glucose was not significantly modified by the peptide. Indeed, the glucose disappearance rate (K of Conard, mg min 10(-2)) changed from 2.5 +/- 0.3 during saline to 2.4 +/- 0.4 during bombesin infusion. 4. When the hypoglycaemic stimulus (i.e. tolbutamide) was used, no effect of the peptide on insulin release could be detected. Here again, the drop in plasma glucose (expressed as Marigo's coefficient) was not affected by the peptide, with a value of 92.8 +/- 12.6 and 84.0 +/- 10.9 during bombesin and saline administration. 5. These data therefore show that, at normal or low blood glucose levels, the dose of bombesin used is unable to modify insulin release and suggest that this peptide might be regarded as a glucose-dependent insulinotropic peptide.