The study examined factors that were associated with outcome in the treatment of PTSD. A trial of cognitive therapy compared to imaginal exposure of chronic PTSD showed that although clinical improvements were obtained after treatment and at 6 month follow-up one type of treatment was not significantly superior to the other. Characteristics of the patient, the trauma and treatment and of pretreatment clinical measures were investigated as predictors of PTSD outcome. Eleven variables were significantly associated with the pre- to post-treatment change in CAPS severity scores. Of these, three (duration of therapy, gender and suicide risk) were selected into a step-wise multiple regression equation to explain 36.5% of the outcome. Similarly, nine variables were significant associated with the pretreatment to follow-up change with three variables (number of missed therapy sessions, residential status and co-morbid GAD) being selected into the equation and explaining 36.9% of the outcome. The best predictor of outcome was inconsistent attendance at therapy.