BACKGROUND: Non-steroidal anti-inflammatory drugs (NSAIDs) induce gastric ulcers. AIMS: To assess whether the somatostatin analogue octreotide prevents NSAID induced mucosal gastrointestinal damage in both animals and humans. The effect of octreotide on neutrophil adhesion to the endothelium was also evaluated. METHODS: Male Sprague-Dawley rats were pretreated either with saline (0.3 ml subcutaneously) or octreotide (0.001-1 ng/kg subcutaneously). After 30 minutes gastric ulcers were induced by the intragastric application of NSAIDs (20 mg/kg indomethacin, 200 mg/kg aspirin, 200 mg/kg ibuprofen, or 50 mg/kg diclofenac). Four hours later the rats were killed and gastric mucosal lesions were assessed by computed planimetry. To determine whether octreotide could prevent indomethacin induced injury in humans, 20 healthy volunteers were evaluated in a double blind, placebo controlled study. RESULTS: Octreotide prevented NSAID induced gastric mucosal lesions (p < 0.05). The dose response curve was U shaped and the most effective dose was 0.1 ng/kg. Leucocyte adherence in submucosal venules of the stomach was evaluated by in vivo microscopy. Octreotide (0.1 ng/kg subcutaneously) prevented indomethacin (20 mg/kg intragastric) induced leucocyte adherence in gastric submucosal venules (p < 0.05). Healthy human volunteers received 50 mg indomethacin orally thrice a day concomitantly with either an identical placebo or 0.01 microgram, 0.1 microgram, or 1 microgram octreotide subcutaneously thrice a day for three days. Injury was assessed by endoscopy. There was a negative correlation between the octreotide dose and injury score (p < 0.03 for gastric injury, p < 0.001 for duodenal injury). CONCLUSIONS: Octreotide protects the stomach from NSAID induced gastric injury, probably via its ability to reduce NSAID induced neutrophilic adhesion to the microvasculature. Octreotide also ameliorated indomethacin induced gastric and duodenal injury in humans.