Gliclazide has been reported to decrease platelet function and to inhibit the progression of diabetic retinopathy in addition to having a hypoglycemic effect. To confirm these effects we performed a double-blind randomized study using glibenclamide as a reference drug. Thirty-eight hospitals from eight university groups in Japan performed the study on type II diabetic subjects. Evaluation of blood glucose control, platelet adhesiveness, platelet aggregation and blood lipids over 24 weeks were assessed by the central committee. Two hundred and eighty-nine patients were enrolled in the study. Twelve were excluded and 277 were statistically analysed. Homogeneity between the two diabetic groups was demonstrated for background factors. Forty mg of gliclazide was comparable to 2.5 mg of glibenclamide in the potency of hypoglycemic efficacy. Funduscopic aggravations were observed in a statistically smaller number of cases in the gliclazide group than in the glibenclamide group and in evaluation of serum lipids, the gliclazide group was also superior to the glibenclamide group. No significant difference between the two groups was found in platelet adhesiveness and aggregation. Gliclazide is a useful drug in the therapy of diabetes mellitus.