Cocaine treatment trials suffer from a high rate of attrition. We examined pretreatment neurophysiological factors to identify participants at greatest risk. Twenty-five participants were divided into concordant and discordant groups following electroencephalogram (EEG) measures recorded prior to a double-blind, placebo-controlled treatment trial. Three possible outcomes were examined: successful completion, dropout, and removal. Concordant (high perfusion correlate) participants had an 85% rate of successful completion, while discordant participants had a 15% rate of successful completion. Twenty-five percent of dropouts and 50% of participants removed were discordant (low perfusion correlate), while only 25% of those who completed were discordant. Failure to complete the trial was not explained by depression, craving, benzoylecgonine levels or quantitative electroencephalogram (QEEG) power; thus cordance may help identify attrition risk.