Objective: To determine the sensitivity of multifocal visual evoked potentials (mVEP) in optic neuritis of an inflammatory or demyelinating nature. Design: Cross-sectional study. Participants: Sixty-four patients participated who had a confirmed diagnosis of optic neuritis (ON) (past and acute). Based on the McDonald multiple sclerosis (MS) criteria, 25 patients (27 eyes with ON) were deemed to have isolated optic neuritis and thus not have MS (i.e., the not-MS group), and 19 patients (24 eyes with ON) had a diagnosis of MS (i.e., the MS group). The remaining 20 patients (25 eyes with ON) were at a high risk of MS, but diagnostic evaluation was equivocal, and thus were classified as the possible MS group. A control group of 20 normal patients was enrolled. Testing: The mVEP test was performed using the Accumap. All ON patients had recent magnetic resonance imaging scans of the brain and spinal cord. Main Outcome Measures: Multifocal visual evoked potentials amplitude and latency values were analyzed within each group and were compared with the normal controls. Results: No abnormality was recorded on mVEP in the control group. Of all the ON eyes, 74 (97.3%) were abnormal on mVEP testing. Amplitude values were abnormal in 92.6% of not-MS eyes, 92.0% of possible MS eyes, and 100% of those with MS, and latency was abnormal in 33.3%, 76.0%, and 100%, respectively. There was a significant difference in the mVEP latency z-scores among all ON groups (P<0.01; Kruskal-Wallis test). Although distribution graphs of latency z-scores in the not-MS and MS groups had single peaks and were clearly separate from each other, the latency z-score distribution within the possible MS group in postacute patients was bimodal, with each peak corresponding to the distribution of the not-MS and MS group, respectively. The mVEP latency z-scores had a sensitivity and specificity of 100% in detecting patients with ON due to MS when compared with normal patients. Conclusions: The mVEP test is a sensitive and specific tool for detecting optic neuritis. There was a significant difference in latency analysis findings between patient groups as classified according to the McDonald MS criteria. Latency results suggest a role in identifying a patient's risk for future MS.