We compared the diagnostic usefulness of evoked potential (EP) studies and magnetic resonance imaging (MRI) in the evaluation of 27 patients with definite or probable multiple sclerosis (MS). MRI scans demonstrated multiple lesions in 21 patients whereas EP studies showed multiple abnormalities in 14 patients (4 of whom had only somatosensory EP abnormalities). Eighteen patients had similar MRI and EP results (e.g., normal or multiple abnormalities), 8 had multiple abnormalities shown by MRI but normal or single-modality abnormal EPs, and 1 had multiple abnormal EPs but a normal MRI scan. There was no significant difference in the sensitivity of the two techniques in detecting multiple lesions in the patients with definite MS, whereas among those with probable MS, MRI had a significantly higher yield. Seventeen patients showed clinical evidence of posterior fossa involvement, 6 patients had abnormal brainstem auditory evoked potentials (BAEPs), and 4 patients had areas of increased signal intensity revealed by MRI of the brainstem. There was no clinical evidence of brainstem involvement in 2 patients with BAEP abnormalities, 2 with an abnormal posterior fossa shown by MRI, and one patient with abnormalities shown by both BAEP and MRI. We conclude that MRI is more sensitive in detecting multiple lesions than are multimodality EP studies, but that BAEP assessment may be slightly more sensitive than MRI in detecting brainstem lesions.