Background: A diagnosis of post-traumatic stress disorder (PTSD) requires the identification of one or more traumatic events, designated the index trauma, which serves as the basis for assessment of severity of PTSD. In patients who have experienced more than one traumatic event, severity may depend on the exact definition of the index trauma. Defining the index trauma as the worst single incident may result in PTSD severity scores that differ from what would be seen if the index trauma included multiple events.Objective: This study aimed to investigate the impact of the definition of the index trauma on PTSD baseline severity scores and treatment outcome.Method: A planned secondary analysis was performed on data from a subset (N = 58) of patients enrolled in a trial evaluating the efficacy of a 12 week residential dialectical behavioural therapy programme for PTSD related to childhood abuse (DBT-PTSD). Assessments of the severity of PTSD were conducted at admission, at the end of the 12 week treatment period, and at 6 and 12 weeks post-treatment, using the Clinician-Administered PTSD Scale. The index trauma was defined with respect to both the worst single incident and up to three qualitatively distinct traumatic events.Results: When the index trauma included multiple traumas, PTSD severity scores were significantly higher and improvements from pre- to post-treatment were significantly lower than when the index trauma was defined as the worst single incident.Conclusions: In patients with PTSD who have experienced multiple traumas, defining the index trauma as the worst single incident may miss some aspects of clinically relevant symptomatology, thereby leading to a possibly biased interpretation of treatment effects. In DBT-PTSD, treatment effects were lower when the index trauma included multiple traumatic events. More research is needed to determine the impact of the various index trauma definitions on the evaluation of other trauma-focused treatments.