The present study investigated predictors of treatment outcome and dropout in two samples of PTSD-patients with mixed traumas treated using prolonged imaginal exposure. Possible predictors were analysed in both samples separately, in order to replicate in one sample findings found in the other. The only stable finding across the two groups was that patients who showed more PTSD-symptoms at pre-treatment, showed more PTSD-symptoms at post-treatment and follow-up. Indications were found that benzodiazepine use was related to both treatment outcome and dropout, and alcohol use to dropout. Demographic variables, depression and general anxiety, personality, trauma characteristics, feelings of anger, guilt, and shame and nonspecific variables regarding therapy were not related to either treatment outcome or dropout, disconfirming generally held beliefs about these factors as contra-indications for exposure therapy. It is concluded that it is difficult to use pre-treatment variables as a powerful and reliable tool for predicting treatment outcome or dropout. Clinically seen, it is therefore argued that exclusion of PTSD-patients from prolonged exposure treatment on the basis of pre-treatment characteristics is not justified.