We investigated the effect of the calcium channel-blocking agent nimodipine on regional cerebral metabolic rate of glucose in acute ischemic middle cerebral artery infarction diagnosed clinically and by computed tomography. Twenty-seven patients entered the study within 48 hours after onset of symptoms and randomly received either nimodipine (2 mg/hr constant intravenous infusion for 5 days, 120 mg/day orally for another 16 days) or placebo. Four of the 27 patients died within the first 3 weeks and could not be evaluated. Of the remaining 23 patients, 11 were assigned to the nimodipine group and 12 to a control group. We analyzed data from positron emission tomography, performed twice before and after completion of therapy, and clinical data from the treatment period and the next 6 months based on the Mathew Score for early assessment and the Barthel Index for late assessment. During the post-treatment period, two patients from the nimodipine group and three from the control group died. The evaluated patients were comparable for age and sex distribution, initial clinical deficit, and infarct size and localization. We found significant metabolic changes between both treatment groups for contralateral infarct mirror region, ipsilateral and contralateral cerebral gray matter, and contralateral and ipsilateral cerebellar hemispheres (side x region x treatment interaction p less than 0.025). The nimodipine group had bilaterally increased regional cerebral metabolic rate of glucose of morphologically intact cerebral (14.6% and 17.1%, respectively) and cerebellar structures (6.9% and 10%, respectively).(ABSTRACT TRUNCATED AT 250 WORDS)