Our aims were to evaluate the prognostic usefulness of magnetic motor evoked potentials (MMEPs) in ischemic stroke, to study the evolution of MMEP abnormalities and the relationships between MMEP abnormalities and infarction topography. We prospectively analyzed 50 consecutive ischemic stroke patients who were followed up to 1 year. MMEPs were recorded 1, 3, 30 and 90 days after stroke and we measured amplitudes and latencies/central motor conduction times (CMCTs) of MMEPs from hypothenar, biceps brachiallis, gastrocnemius and quadriceps. Univariate and multivariate analyses of the clinical and MMEPs data were performed. Patients with Rankin 0-3 at 1 year had had acutely MMEPs with shorter latencies and higher amplitudes than patients with Rankin 4-5 or deceased patients. Increased blood pressure correlated with increased survival, whereas increased heart rate and hyperglycemia correlated with increased mortality. The variables infarction size on second CT, age, and first day CMCT-S1 correctly classified 1 year outcome on discriminant analysis. The inclusion of MMEPs values increased the probability of correct classification from 76% to 84%. We conclude that in patients with nondisabling strokes MMEPs may have an independent value in the prediction of prognosis, increasing the accuracy of prognosis calculations made employing clinical and laboratory data. Topography of lesions should be considered when analyzing MMEP abnormalities after stroke.