Thirty-three patients with comorbid PTSD and cocaine dependence were divided into two groups: one in which the trauma and PTSD occurred before the onset of cocaine dependence (primary PTSD) and one in which the PTSD occurred after cocaine dependence was established (primary cocaine). In the primary-PTSD group, the trauma was generally childhood abuse. In the primary-cocaine group, the trauma was generally associated with the procurement and use of cocaine. In the primary-PTSD group, there were significantly more women, more other Axis I diagnoses, more Cluster B and C Axis II diagnoses, and more benzodiazepine and opiate use. In the primary-cocaine group, there was a trend toward more cocaine use in the previous month. Significant clinical differences between these two groups may warrant different types of treatment or differing treatment emphasis. The authors caution that this is a preliminary study that is limited by small sample size and the retrospective nature of the data collected. Further, the fact that all of the subjects were recruited from a pharmacological treatment trial may introduce some selection bias. 4 tables and 24 references