Cortical glucose utilization was measured under light anaesthesia using positron emission tomography and 18F-fluorodeoxyglucose before and serially after stereotaxic anterior corpus callosum section in 3 adult baboons (Papio papio); in 1 animal needles were introduced but callosotomy was not performed ('sham' operation). Lesion efficacy was verified by x-ray computerized tomography and by postmortem examination which indicated effective anterior two-thirds callosal section in all 3 baboons, as well as a small contusional area in the medial frontal cortex. A two-way layout analysis of variance performed on the regional cerebral metabolic rate of glucose (CMRGlc) data obtained in the frontotemporal, temporal and occipital areas revealed significant effects for both the time and the region factors (P less than 0.00001 for each); the time effects indicated a significant initial metabolic depression (P less than 0.01) at day 11 and 18, and a subsequent recovery (P less than 0.01 to 0.05) at day 98. Average CMRGlc changes from preoperative values at successive studies were -23%, -27%, -17%, -8% and +2%, and -17%, -21%, -10%, -5% and +6% in the 'surgically approached' and the contralateral frontotemporal area, respectively. The frontotemporal/occipital CMRGlc ratio was significantly depressed at days 11, 18 and 32 (P less than 0.01, 0.01 and 0.05, respectively), indicating a transient postoperative alteration in the cortical metabolic pattern. In the 'sham' animal, surgery induced negligible metabolic effects. Our results indicate that effective anterior corpus callosum section induces a transient bilateral depression of cortical metabolism, which predominates in the anterior parts of the brain, while metabolic recovery suggests efficient neuronal adaptation. These functional effects in the cortex may underlie the transient behavioural changes that follow anterior callosotomy in man.