The goal of this study was to test the hypothesis that the synthesis/release of hydroxyl radical accounts for impaired nitric oxide synthase-dependent dilatation of the basilar artery during diabetes mellitus. We measured the diameter of the basilar artery in vivo in nondiabetic and diabetic rats (streptozotocin, 50-60 mg/kg ip) in response to nitric oxide synthase-dependent agonists (acetylcholine and substance P) and a nitric oxide synthase-independent agonist (nitroglycerin). Reactivity of the basilar artery was measured in untreated nondiabetic and diabetic rats and in nondiabetic and diabetic rats treated with a daily intraperitoneal injection of dimethylthiourea (DMTU; 50 mg/kg). Injection of DMTU was started 48 h after injection of streptozotocin and was continued throughout the diabetic period (3-4 wk). Topical application of acetylcholine (0.1, 1.0, and 10 microM) and substance P (0.1 and 1.0 microM) produced similar dilatation of the basilar artery in untreated and DMTU-treated nondiabetic rats. In untreated diabetic rats, the magnitude of vasodilation produced by acetylcholine and substance P was significantly less than in untreated nondiabetic rats. However, in DMTU-treated diabetic rats, dilatation of the basilar artery in response to acetylcholine and substance P was similar to that observed in nondiabetic rats. Dilatation of the basilar artery in response to nitroglycerin was similar in untreated and DMTU-treated nondiabetic and diabetic rats. These findings suggest that impaired nitric oxide synthase-dependent dilatation of the basilar artery during diabetes mellitus may be related to the synthesis/release of hydroxyl radical.