Background Models of intraluminal middle cerebral artery occlusion present an intrinsic variability in infarct size. Behavioral evaluation is frequently performed during arterial occlusion to confirm success of surgery. Aims and/or hypothesis We compared the value of behavioral testing and multimodal magnetic resonance imaging performed during arterial occlusion for identifying successfully operated animals. Methods Rats were tested with behavioral assessment (using three scoring scales and the adhesive removal test) and multimodal magnetic resonance imaging (including magnetic resonance angiography, diffusion-weighted and perfusion-weighted imaging), both performed during the two-hours of middle cerebral artery occlusion using the intraluminal suture model. Behavioral assessment was repeated 24 h after reperfusion, followed by sacrifice. Results Acute apparent diffusion coefficient lesion volume was correlated with both 2,3,5-triphenyl tetrazolium chloride infarct size ( r = 0.75, P = 0.02) and behavioral status ( r = 0.66, P = 0.05) on day one. Conversely, no correlation was found between acute behavioral examination and day one outcomes (2,3,5-triphenyl tetrazolium chloride infarct volume, r = 0.40, P = 0.28; behavioral examination, r = 0.39, P = 0.30). Day zero apparent diffusion coefficient volumes ( P = 0.04), but not behavioral assessment ( P = 0.60), discriminated animals with day one corticostriatal infarcts from these with subcortical infarcts. Conclusions Acute behavioral testing performed during arterial occlusion fails to identify successfully operated animals. Acute diffusion magnetic resonance imaging may be more appropriate to assess and reduce infarct size variability in this model.