The mechanism by which halothane, isoflurane, and nitrous oxide increase cerebral blood flow (CBF) is unknown. We assessed the cerebrovascular effects of nitrous oxide (70%; n = 6), isoflurane (1 minimum alveolar anesthetic concentration: 1.4%; n = 6) or halothane (1 minimum alveolar anesthetic concentration: 0.8%; n = 6) before and after blockade of nitric oxide (NO) synthase with 40 mg/kg Nω-nitro-L-arginine methyl ester (L-NAME) intravenously in dogs with baseline pentobarbital anesthesia. Baseline CBF (microspheres) was determined after 1 h of pentobarbital anesthesia. Cerebral perfusion pressure (CPP) was maintained during inhaled anesthetic or L-NAME by either hemorrhage or inflation of an intra-aortic balloon. Before L-NAME, halothane and isoflurane increased CBF (40 ± 4 to 56 ± 6 mL·min−1· 100 g−1 and 43 ± 6 to 78 ± 12 mL·min−1·100 g−1, respectively) with no change in cerebral oxygen consumption (baseline: halothane, 2.6 ± 0.2; isoflurane, 2.0 ± 0.2 mL·min-1·100 g−1). On the contrary, nitrous oxide increased CBF similarly (40 ±6 to 57 ± 8 mL·min−1·100 g−1), but increased cerebral oxygen consumption (2.2 ± 0.3 to 3.0 ± 0.3 mL·min−1· 100 g−1). L-NAME decreased blood flow in the neurohypophysis by 80% with no change in blood flow in other brain regions. After L-NAME, reexposure to nitrous oxide, halothane, or isoflurane resulted in no change in CBF. These data demonstrate that blockade of NO synthase by L-NAME prevents increased CBF by nitrous oxide, isoflurane, and halothane and suggests that the mechanism of increased CBF with all three anesthetics is mediated via a NO or nitrosothiol mechanism.