Research has demonstrated a cyclical relationship between posttraumatic stress disorder (PTSD) and substance use disorder. Identifying factors that link PTSD symptom clusters and substance use disorder may illuminate mechanisms underlying the PTSD-substance use disorder relationship, better informing interventions that target this comorbidity. The current study of individuals enrolled in an outpatient aftercare chemical dependency program in King County, Washington, assessed whether overall PTSD symptoms and specific PTSD symptom clusters predicted craving depending on individuals identified primary drug of choice (DOC). Participants eligible for the parent study were at least 18 years of age, fluent in English, medically cleared from substance withdrawal, and able to participate in treatment sessions and agreed to random assignment. Random assignment to either a mindfulness-based relapse prevention group, a standard relapse prevention group, or a treatment as usual group was conducted on a computer randomization program. A secondary analysis of baseline data was employed in the current study to determine which of the PTSD symptom clusters (avoidance, hyperarousal, and intrusion) predicted substance craving. Covarying for severity of dependence, results suggest that overall PTSD scores predicted craving in participants who identified alcohol, stimulants, and opiates as their primary DOC. Further, avoidance-related PTSD symptoms alone predicted a significant proportion of the variability in craving in stimulant users, and hyperarousal symptoms alone predicted a significant proportion of the variability in craving in alcohol users. No specific PTSD cluster significantly predicted a proportion of the variability in craving in marijuana or opiates users. Findings suggest that craving may play a role in maintaining the relationship between specific PTSD symptom clusters and substance use disorder, and the nature of this relationship may differ by primary DOC. The clinical trial on which this secondary analysis of data was conducted is registered as NCT01159535 at www.clinicaltrials.gov.The original trial from which data for this study was drawn was supported by the National Institutes of Health [NIH/NIDA 5 R01 DA025764-02].