BACKGROUND AND OBJECTIVE Little is known about direct comparisons of the efficacy of trauma-focused psychotherapies and SSRIs. This is the first randomized clinical trial comparing the efficacy of prolonged exposure (PE), paroxetine (Ph) and their combination (Comb) in a sample of adults diagnosed with PTSD following motor vehicle accidents (MVA). METHODS A total of 228 people were randomly assigned to a twelve-week treatment of PE (N = 114), Ph (N = 57) or Comb (N = 57). RESULTS The ITT analyses showed that the remission rate of PTSD was significantly greater after PE (65.5%) compared with Ph (43.3%), whereas Comb (51.2%) did not differ from either. The differences in dropout rates were not significant between treatments (18.4% - PE; 12.2% - Ph; 22.8% - Comb), while the differences in numbers of refusers were significant (3.5% PE <31.6% Comb <47.4% Ph; p < .01). The changes in self-rated PTSD were significant for each treatment and without significant differences between treatments. Differences between clinician and self-rated outcomes can be explained by depressive symptoms influencing self-rating by the PDS. At a 12 month follow-up treatment results were maintained and different trajectories of functioning were identified. LIMITATIONS Larger samples would allow analyses of predictors of treatment response, dropout and refusal. CONCLUSIONS In this, largest to date study comparing PE, paroxetine and combination treatment in PTSD PE was more effective than Ph in achieving remission of PTSD. The additive effect of Comb over any monotherapy was not shown.