In vitro studies have demonstrated that gliclazide has free radical scavenging and antiplatelet activities. To assess this clinically, we studied gliclazide in a blinded, randomized, glibenclamide-controlled trial in 30 type II diabetic patients with retinopathy. All patients had been taking glibenclamide for more than 12 months before being randomized to receive either an equipotent dose of gliclazide or to continue on glibenclamide. Diabetic control was not modified. The patients were well matched at randomization (mean age, 58 years; duration of diabetes, 8 years; 20 males; mean hemoglobin A1 [HbA1], 8.6%) and their degree of diabetic control was not altered during the trial. Free radical activity was assessed as oxidative status by plasma thiols (PSH), lipid peroxides (MDA-LM), and red blood cell superoxide dismutase activity (SOD). Platelet aggregation in whole blood to collagen (Plt-ag) was used as the measure of platelet reactivity. There were no differences between these measurements at baseline. At 3 months, the oxidative status and platelet aggregation in the gliclazide group had improved significantly compared with baseline and had also showed significant differences in all parameters when compared with the glibenclamide group. Therefore, comparing gliclazide with glibenclamide-treated patients at 3 months, we found: PSH, 458 ± 38 versus 414 ± 34 µmol/L, P < .004; MDA-LM, 7.0 ± 0.6 versus 8.3 ± 0.8 µmol/L, P < .0002; SOD, 152 ± 36 versus 123 ± 15 µg/mL, P < .016; Plt-ag, 50.8 ± 24 versus 72.3% ± 15%, P < .006. These changes were maintained at 6 months. These results confirm in a clinical study that gliclazide is a powerful general free radical scavenger. The effects on platelets may be secondary to this, and the free radical scavenger effect appears to be independent of diabetic control.