Background and Purpose-The effects of acute smoking on cerebral circulation are controversial. This study was designed(1) to clarify any differences between the effects of cigarette smoking and nicotine infusion and between the effects of single-and multiple-cigarette smoking on cerebral vessels and (2) to probe the mechanism(s) underlying the vascular responses. Methods-In pentobarbital-anesthetized, mechanically ventilated Sprague-Dawley rats, pial vessel diameters were measured with the use of a cranial window preparation. We studied the effects of (1) 60 puffs per minute of mainstream cigarette smoke from cigarettes having 2 nicotine levels (0.1 and 1 mg per cigarette), (2) administration of nicotine (0.05 mg per body IV), and (3) repeated smoking (four 1 mg nicotine-containing cigarettes at 30-minute intervals) (nϭ6 each). Results-Inhalation of smoke from a 0.1 or 1 mg nicotine-containing cigarette for 1 minute caused pial arterioles to constrict at 30 seconds (7.2% and 7.3%, respectively) and then to dilate (peak at 5 to 10 minutes; 4.6% and 17.9%, respectively). Nicotine infusion caused pial vasodilation (35.7%) without an initial vasoconstriction. Repeated smoking suppressed the pial vasodilation but not the initial vasoconstriction. The vasodilation induced by a single cigarette was greatly inhibited by pretreatment with mecamylamine or glibenclamide and attenuated by propranolol or N -nitro-Larginine methyl ester; the initial vasoconstriction was inhibited by seratrodast, a thromboxane A 2 receptor antagonist (nϭ6 in each case). Conclusions-Single-cigarette smoking had a significant biphasic effect on cerebral arteriolar tone. The vasodilation was attenuated by repeated smoking. The vasodilation is most likely an effect of nicotine, at least in part mediated via sympathetic activation, NO production, and K ϩ channel activation. The vasoconstriction is partially due to thromboxane A 2 induced by cigarette smoke. (Stroke. 1998;29:1656-1665.)