PURPOSE: To determine whether the analysis of abnormally high signal intensities in ischemic tissue, as revealed by diffusion-weighted MR imaging (DWI) can be used to evaluate reversible brain lesions in a cat model of acute ischemia. MATERIALS AND METHODS: Ten cats were divided into two groups of five (Group I and Group II), and in all animals the middle cerebral artery was temporarily occluded. Group I underwent T2-DWI 30 minutes after occlusion, and Group II 120 minutes after occlusion. In both groups, DWI was performed one hour and 24 hours after reperfusion (at one hour, non-T2-weighted; at 24 hours, T2-weighted). Both occlusion and reperfusion were monitored by 99m TC-ECD brain perfusion SPECT. All animals were sacrificed 24 hours later and their brain tissue was stained with TTC. Signal intensity ratios (SIR, signifying average signal intensity within the region of interest divided by that in the contralateral, nonischemic, homologous region) of the two groups, as seen on DWI were compared. The percentage of hemispheric lesions occurring in the two groups was also compared. RESULTS: SIR after occlusion of the middle cerebral artery was 1.29 in Group I and 1.59 in Group II. Twenty-four hours after reperfusion, SIR in Group I was higher than in Group II (p<0.01). After occlusion and reperfusion, the percentage of hemispheric lesions in Group I was less than in Group II. For the latter, the percentage of these lesions revealed by TTC staining and T2-weighted imaging was 48% and 59%, respectively, findings distinctly different from those for Group I. In addition, in group I, infarction was revealed by neither TTC staining nor T2-weighted imaging (p<0.01). CONCLUSION: The use of DWI to evaluate signal intensity ratios can help determine whether or not brain injury after temporary cerebral ischemia is reversible.