Cocaine-dependence has been shown to affect the amplitudes of the P50 mid-latency auditory evoked response (MLAER) as well as P50 sensory gating. The effects on subsequent MLAERs (N100 and P200) have not been examined. The objective of the current study was to further assess the effects of chronic cocaine use on the P50, N100, and P200 components. Thirty-four, at least three weeks abstinent, cocaine-dependent individuals and 34 age and gender matched healthy controls were examined. The amplitudes, latencies and gating measures were calculated and compared between the groups. The N100 and P200 were significantly smaller in patients as compared to control subjects. Sensory gating of the P50, the N100, and the P200 were deficient in cocaine-dependent subjects. Latencies of all measured components were prolonged in subjects who reported developing paranoia while intoxicated. Finally, a positive correlation was found between length of abstinence and evoked response amplitudes. We conclude that the effects of cocaine on sensory gating extend beyond the P50 to the N100 and the P200 components. The data also suggest that prolonged latency of the evoked potentials may be a correlate of cocaine-induced psychosis. Finally, the data suggest that some recovery of amplitude and gating occurs with abstinence.