This study tested the hypothesis that local cerebral O2 supply and consumption are not precisely balanced and that vascular alpha adrenoceptors affect this equilibrium. Twenty-two male Long-Evans rats were used in this study. Cerebral blood flow was determined in six regions, using 4-iodo [N-methyl-14C]antipyrine. Oxygen saturation was determined microspectrophotometrically in small veins draining three regions of the brain of the rat. Comparisons were made between a control group and a group given N-methyl chlorpromazine, an alpha adrenergic blocker that does not affect central neuronal alpha adrenoceptors. Under control conditions, no regional differences in flow were found. With N-methyl chlorpromazine, cerebral blood flow in the basal ganglia was greater than all other regions, except the frontal cortex. Cerebral O2 consumption was higher in the frontal cortex than the medulla under control conditions and this difference was not statistically significant after N-methyl chlorpromazine. Cerebral venous O2 saturations (a measure of cerebral O2 supply/consumption balance) were found to be significantly heterogeneous under control conditions. The coefficient of variation (CV = 100 x SD/mean) averaged 18%. The average cerebral venous O2 saturation was 59 +/- 11%. Administration of N-methyl chlorpromazine significantly reduced this heterogeneity through a reduction in the number of veins with low O2 saturations (CV = 11%). The average value increased slightly but significantly to 62 +/- 8%. Thus, N-methyl chlorpromazine eliminated many microregions of high O2 extraction. This indicated that vascular alpha adrenoceptors limit cerebral blood flow to some of the brain regions.