Ten patients with diabetic gastroparesis were selected for a randomized, double-blind, controlled trial of metoclopramide. Each patient had longstanding insulin-requiring diabetes mellitus and symptoms of gastric stasis. The patients were evaluated for the symptoms of gastric stasis and radionucleotide gastric emptying was measured before the patients entered the study and after they were given either metoclopramide or placebo treatment. Metoclopramide, 10 mg orally, stimulated an increase in the rate of gastric emptying (56.8% +/- 7.4%) in contrast to the response to placebo (37.6% +/- 7.7%) (p less than 0.01). The overall symptoms and symptoms of vomiting were markedly reduced during metoclopramide treatment in contrast to those during placebo treatment. Before the study five patients were constipated (less than three bowel movements per week); during metoclopramide treatment the patients' bowel habits were improved. There was a poor correlation between improved gastric emptying and decreased symptoms. Metoclopramide may improve symptoms of diabetic gastric stasis through two mechanisms: its peripheral effect on gastric smooth muscle, which increases gastric emptying; and its central effects on the chemoreceptor vomiting zone, which decrease nausea.