The aim of the present study was to assess the effect of post ictal administration of the pyrrolopyrimidine lipid peroxidation inhibitor, U-101033E, on infarct volume and neuronal and astrocytic metabolism in rats with transient middle cerebral artery occlusion (MCAO). Rats were subjected to 120 min of MCAO followed by 140 min of reperfusion and randomly assigned to control (n=17) or U-101033E treatment (n=16). Drug infusion started 5 min after MCAO and lasted 220 min with a 15 min interruption during the reperfusion procedure. Sixteen rats underwent diffusion weighted imaging 260 min after ictus, from which the apparent diffusion coefficient (ADC) was determined. Seventeen rats received an iv bolus injection of [1-13C]glucose and [1,2-13C]acetate 245 min after ictus. Tissue extracts from two brain regions representing penumbra and ischemic core were analyzed with 13C NMRS and HPLC. U-101033E did not affect the volume of ischemic tissue estimated from the ADC maps. In the penumbra, U-101033E specifically decreased mitochondrial pyruvate metabolism via both pyruvate dehydrogenase and pyruvate carboxylase pathways. Thus, U-101033E impaired both neuronal and astrocytic mitochondrial pyruvate metabolism. At the same time anaerobic glucose usage was increased, leading to increased lactate labeling and content. Also alanine labeling was increased. The data do not support lactate as an important substrate for neuronal mitochondria in ischemia-reperfusion. A similar pattern of reduced mitochondrial pyruvate metabolism and increased cytosolic pyruvate metabolism was found in the irreversibly damaged ischemic core. The present study highlights the importance of other outcome measures than ischemic tissue volume for evaluation of drug efficacy in animal models, which in turn could increase the likelihood of success in clinical trials.