To investigate the central nervous system circulation during spinal anesthesia, local spinal cord blood flow (SCBF) and cerebral blood flow (CBF) were measured simultaneously by the hydrogen clearance technique following subarachnoid lidocaine, phenylephrine, or a combination of both. The mean control values of SCBF and CBF were 22.4 ± 7.9 ml.100 g−1·min−1and 53.1 ± 12.0 ml·100 g−1·min−1, respectively, in dogs lightly anesthetized with halothane.The subarachnoid administration of lidocaine solutions (1, 2, 3, and 5%), 1 ml, failed to produce statistically significant changes in SCBF (P > 0.05). Whereas, when phenylephrine (0.1, 0.2, 0.3, and 0.5%), 1 ml, was injected into the spinal subarachnoid space, SCBF decreased significantly with concentrations greater than 0.2% (P < 0.05). When a mixture of lidocaine (24 mg) and phenylephrine (1 mg) was administered into the subarachnoid space, SCBF decreased significantly and returned to control within 60–90 min. CBF did not change significantly with any of the injections, remaining within less than ±12% of control. Dextrose solutions in water (5 and 7.5%), which were used for dilution of the drugs, did not affect either SCBF or CBF. These results indicate that local spinal cord blood flow can be affected significantly during spinal anesthesia when phenylephrine is added to the local anesthetic solution. However, the circulatory effects of drugs injected into the spinal subarachnoid space appear to be restricted to the local spinal cordper seand do not involve other parts of the CNS.