The utility of Andersen's (1993, 1994) model of psychologic morbidity following cancer treatment for predicting PTSD symptoms in breast cancer survivors (N = 82) was examined. PTSD symptoms, physical comorbidity, social support, depression history, and pre-cancer traumatic stressors were assessed in a structured telephone interview. Multiple regression analysis indicated that Andersen model variables (physical comorbidity, education, disease stage, cancer treatment, depression history, social support) accounted for 39% of variance in PTSD symptom reports (p < .001). Addition to the model of time since treatment completion, pre-cancer traumatic stressors, age at diagnosis, and tamoxifen usage accounted for an additional 16% of variance (p < .001). Higher levels of PTSD symptoms were associated with less social support, greater pre-cancer trauma history, less time since treatment completion, and more advanced disease.