Animal studies have shown that a course of electroconvulsive shock (ECS) leads to a significant reduction in glucose metabolism in rat brains 1 day after the last ECS. In humans, of the two positron emission tomography (PET) studies that assessed the effects of a course of electroconvulsive therapy (ECT) on brain glucose metabolism in depressed patients, one reported no change while the other found a trend for reduction in glucose metabolism in frontal cortical region 24 hours after last ECT. The changes in glucose metabolism detected 24 hours after the last ECS/ECT treatment might simply be due to subacute effects of a seizure. We hypothesized that the changes in brain metabolism that persist 1 week after a course of ECT are more likely to underlie the therapeutic effects of ECT. We, therefore, investigated the effects of a course of ECT on brain glucose metabolism 1 week after last ECT by using PET and [18F]fluorodeoxyglucose (FDG). Six patients who met DSM-IV criteria for a diagnosis of major depressive disorder (unipolar), and were referred for ECT as the clinically indicated treatment were recruited. They underwent two PET scans, one prior to first ECT and the second a week after last ECT. The number of ECT treatments subjects received ranged from 8 to 12 with a mean of 11. Five out of six patients responded to the ECT treatment. Cerebral metabolic rates for glucose were slightly lower in most regions post treatment compared with pretreatment but the differences were not statistically significant. Similarly, there was no significant correlation between changes in regional cerebral metabolic rates for glucose (rCMRglc) and changes in Hamilton Depression Rating Scale (HAM-D 21-item) scores. Our results might suggest that rCMRglc rates are not altered 1 week after a therapeutic course of ECT in depressed patients. Further studies using new generation PET scanners, which have a higher resolution, in larger numbers of depressed patients, are clearly needed before firm conclusions can be drawn.