The authors investigated the spinal blood flow and metabolic effects of subarachnoid clonidine in conscious rats prepared with chronically implanted subarachnoid catheters. For the blood flow experiments, rats received saline (n = 7) or clonidine 20 nmol (7 μg; n = 6), 100 nmol (27 μg; n = 5), or 400 nmol (107 μg, n = 7) intrathecally. Another group of rats received clonidine 400 nmol intravenously (n = 4). Spinal glucose utilization was measured in rats that received either saline (n = 5) or clonidine 100 nmol (n = 5) intrathecally. Spinal cord blood flow (SCBF) and glucose utilization were measured in five gray and three white matter areas of lumbar spinal cord 15 min after drug administration with the autoradiographic iodo-[14C]-antipyrine and 2-[14C]-deoxyglucose methods, respectively. Physiologic differences between the groups were minor. Rats in the blood flow experiments that received clonidine 100 nmol had a slightly lower arterial P(O2) level (70 ± 1 vs. 82 ± 3 mmHg; P < 0.05), whereas those in the glucose utilization group were mildly hypobaric (P(CO2) 27 ± 1 vs. 32 ± 2 mmHg; P < 0.01) relative to control animals. Only animals that received 400 nmol clonidine intrathecally had significant analgesia, as assessed by the tail-flick test. One control animal for the metabolism experiments was technically unsatisfactory and was excluded from data analysis. Subarachnoid clonidine reduced both SCBF and glucose utilization. In spinal gray matter, the largest decreases in flow (32-44%; P < 0.01) occurred with 20 nmol clonidine, whereas flow decreased least (12-27%) with the 400-nmol dose. On the other hand, white matter blood flow decreased (17-39%) only at the two higher doses. Intravenous clonidine 400 nmol reduced SCBF to the same extent as the same dose administered intrathecally. At the single dose studied, subarachnoid clonidine reduced glucose utilization 11-42% in spinal gray and white matter, although the changes were significant in only five of eight areas. Neither the blood flow nor the glucose utilization changes were confined to areas of spinal cord known to contain high concentrations of alpha-2 receptors. These results indicate that, even at subanalgesic doses, subarachnoid clonidine reduces SCBF substantially and that the decrease is associated with, and perhaps caused partially by, a decrease in spinal metabolic rate. Accordingly, these data suggest that the blood flow and analgesic effects of clonidine occur by different mechanisms and support the hypothesis that spinal vasoconstriction may contribute to subarachnoid clonidine's ability to prolong spinal anesthesia.