We examined whether metoclopramide would improve the success rate of transpyloric intubation of a weighted Corpak feeding tube when fluoroscopic guidance is not used. Seventy patients were randomized in a prospective, double-blind fashion to receive either placebo (n = 35) or metoclopramide, 10 mg (n = 35) parenterally, administered immediately after the feeding tube was inserted. Tube location was determined independently by two observers who examined radiographs obtained after barium was instilled via the tube. There was no significant increase in the success rate of duodenal intubation in the total group following metoclopramide, 60%, compared to placebo, 49%. However, analysis of subgroups among the placebo-treated patients revealed that diabetes mellitus, but not other medical conditions, decreased the success rate for duodenal intubation, 20 vs 60% (p less than 0.05). Among diabetic patients, metoclopramide resulted in a significant increase in duodenal placement compared to placebo (p less than 0.05). We conclude that parenteral metoclopramide significantly increases the frequency of transpyloric intubation with small feeding tubes without fluoroscopic guidance in diabetic patients but not in nondiabetic patients.