The present pilot study investigated whether acceleration of gastric emptying in patients with type 1 diabetes and delayed gastric emptying (a possible cause of poorly controlled diabetes) improves long-term glucose control. Eight outpatients with diabetes (age 28-63 years, mean diabetes duration 24.6+/-6.0 years) and delayed gastric emptying of radio-opaque markers were randomised and treated, for three months each, with a prokinetic drug (cisapride 20 mg twice daily) and placebo. Mean capillary glucose, glucose variability (M-values, MAGE), fructosamine, and HbA1c were assessed. Gastric emptying of a solid standard meal was measured by scintigraphy after each treatment period. Chronic administration of a prokinetic drug resulted in improved solid gastric emptying (percentage residual) at 120 min (p=0.025). The percentage residual was 43.6+/-9.6 % during prokinetic treatment and 59.7+/-9.9 % during placebo (standard error of paired differences 5.7 %). The mean gastric emptying time (t/2) of solids was 88 min during prokinetic treatment compared to 113 min in the placebo arm (SE of paired differences 14 min; p=0.09). Mean blood glucose values (9.0+/-3.8 vs. 8.8+/-3.7 mmol/l), daily glucose variability (MAGE 6.8+/-1.3 vs. 6.3+/-1.6 mmol/l; M-value 15.2+/-2.5 vs. 13.9+/-4.5), and HbA1c at 3 months (7.8+/-1.1 % vs. 7.6+/-1.0 %) were not statistically different between prokinetic drug and placebo treatment. Similarly, the frequency of hypoglycaemic episodes (< or = 3 mmol/l) was not different in both groups (78 vs. 68). Our pilot study showed that long-term acceleration of gastric emptying had no effect on overall glycaemic control, the magnitude of glucose excursions, and hypoglycaemic episodes in patients with diabetic gastroparesis. We do not recommend, therefore, acceleration of gastric emptying as treatment strategy for "brittle diabetes" in patients with type 1 diabetes.