Ghrelin has been shown to accelerate gastric emptying in animals where its effect appeared mediated through the vagus nerve. We aimed to verify the gastrokinetic capacity of ghrelin in human. Patients with gastroparesis attributed to a neural dysregulation by diabetes (n = 5) or surgical vagotomy (n = 1) were evaluated. The emptying of a test meal (420 kcal) was determined by the C13 octanoic acid breath test. Saline or synthetic ghrelin 1-4 microg/kg were given in 1 min bolus at the end of the meal. T-lag and T-1/2 were shorter during ghrelin than during saline administration [33 +/- 5 min versus 65 +/- 14 min (p < 0.01) and 119 +/- 6 min versus 173 +/- 38 min (p < 0.001)]. Ghrelin injection therefore accelerated gastric emptying of a meal in humans even in presence of a deficient gastric innervation.