BACKGROUND Mannitol has been used in routine neurosurgical practice for the control of increased intracranial pressure. The effect of mannitol on focal cerebral ischemia was evaluated by somatosensory-evoked potentials (SEP) and magnetic resonance imaging (MRI). METHODS The left middle cerebral artery (MCA) was exposed via the superomedial transorbital approach and occluded proximal to the origin of the perforating arteries. Ten cats received mannitol (0.5 g/kg IV) immediately, 6, 12, and 18 hours after MCA occlusion. The other 10 cats received saline solution and served as control. The animals were initially prepared to measure SEP before and 15, 30, and 60 minutes after MCA occlusion. Following SEP measurement, all cats were prepared for MRI. Sequential MRI of both intravoxel incoherent motion (IVIM) and T2-weighted spin echo techniques were obtained at 2, 4, 6, and 24 hours after MCA occlusion. The animals were sacrificed after the last MRIs for histologic study. RESULTS The SEP amplitude decreased to about 10% at 15 minutes after MCA occlusion and then gradually recovered to 38% at 60 minutes in the mannitol group, and 21% in the control group. In MRI study, IVIM imaging demonstrated ischemic cerebral injury as a sharply demarcated area at 2 hours after MCA occlusion, while T-2 weighted imaging failed to show clear evidence of injury until 2-6 hours. High-signal intensity areas on both IVIM and T2-weighted images were smaller in the mannitol group than those in the control group. Histologic study demonstrated that infarction size was 36.9% +/- 7.7% of the left hemisphere in the mannitol group and 57.3% +/- 5.3% in the control group (p < 0.05). CONCLUSIONS Mannitol is effective for acute cerebral ischemia, and SEP and MRI are useful for monitoring it.