Glibenclamide inhibits the opening of vascular ATP-sensitive potassium (K(ATP)) channels, which represents a protective mechanism during ischaemia. This effect may imply harmful cardiovascular effects of glibenclamide when used under conditions of ischaemia in patients with Type II diabetes. Acarbose is not associated with effects on the cardiovascular system, because the drug is not absorbed from the bowel. Therefore we hypothesized that treatment of Type II diabetes patients with glibenclamide will impair the vasodilator function of K(ATP) opening, unlike treatment with acarbose. A double-blind randomized cross-over study in 12 patients with Type II diabetes was performed to compare the effects of glibenclamide with those of acarbose on the vasodilator responses to K(ATP) channel opening in the forearm vascular bed. The study consisted of two periods: 8 weeks of treatment with orally administered glibenclamide (10 mg x day(-1)) followed by 8 weeks of treatment with acarbose (300 mg x day(-1)), or vice versa. At the end of each treatment period, forearm blood flow (venous occlusion plethysmography) in response to intra-arterially administered diazoxide, acetylcholine and dipyridamole and to forearm ischaemia was measured. The diazoxide-mediated increase in the forearm blood flow ratio (infused/control arm) was significantly less pronounced after glibenclamide than after acarbose (290 +/- 58% and 561 +/- 101% respectively; P<0.0005). Forearm blood flow responses to acetylcholine, dipyridamole and forearm ischaemia were similar during glibenclamide and acarbose treatment. Thus, in patients with Type II diabetes mellitus, treatment with glibenclamide is associated with an attenuated response to K(ATP) opening as compared with treatment with acarbose. This implies that glibenclamide may affect defensive mechanisms under conditions of K(ATP) channel activation.