The auditory P50 evoked response to click stimuli was recorded from 10 2-week abstinent African-American chronic cocaine abusers and 10 African-American non-substance-abusing controls. Stimuli consisted of pairs of clicks with a 500-msec interval between clicks in a pair, and a 7-8 sec interval between pairs of clicks. After averaging responses to 100 pairs of clicks and digital bandpass filtering between 10 and 50 Hz, P50 amplitude to the first and the second click was measured. The conditioning/testing (C/T) ratio, an index of P50 suppression, was computed as the ratio of P50 amplitude to the second compared to the first click. Chronic cocaine abusers had markedly diminished P50 amplitudes and increased C/T ratios (indicating decreased P50 suppression) in comparison to the controls. These P50 abnormalities were not seen in additional Caucasian gay/bisexual comparison groups of active alcoholics (n = 15) and non-substance-abusing controls (n = 10). Thus, decrements in P50 amplitude and P50 suppression appear to be specific to cocaine abuse, and to differentiate cocaine abuse from alcohol abuse. A response analogous to P50 can be measured in animals, facilitating the development of animal models of this cocaine effect.