Objectives: The goal of the present study was to evaluate the relative effects of cocaine, cocaine and alcohol, or opioid dependence on P300 event-related potentials (ERPs). In addition, the effects of selected premorbid and comorbid factors were examined. Methods: P300 ERPs were recorded from 72 residential treatment program patients, characterized by a history of either cocaine (n = 25), or cocaine and alcohol (n = 18), or opioid (n = 29) dependence, and 14 non-drug-dependent community volunteers. The 86 subjects completed a visual continuous performance test formed by a series of 150 presentations of individual consonant letters. They were asked to press a key whenever a letter was presented twice in succession. Results: Analyses of P300 ERPs obtained on target trials revealed a similar amplitude decrement in all the patient groups. Further analyses of P300 activity in the 3 drug-dependent patient groups revealed a negative correlation between P300 amplitude and the number of DSM-M childhood conduct disorder criterion behaviors as well as a positive correlation between P300 amplitude and the duration of drug abstinence. Conclusions: P300 amplitude in drug-dependent patients is influenced by a complex interaction between CNS pathology that predates, and probably promotes, the onset of drug dependence and CNS pathology that resolves during the process of recovery from drug dependence. (C) 2001 Elsevier Science Ireland Ltd. All rights reserved.