Objective: Childhood sexual abuse (CSA) increases risk for posttraumatic stress disorder (PTSD), and both CSA and PTSD are strongly associated with impaired social functioning. Interpersonal psychotherapy (IPT) has demonstrated promise for adults with PTSD and has been shown to be effective for women with PTSD and histories of CSA, but the mechanisms by which IPT might improve PTSD are unknown. The current secondary analysis tested whether the ameliorative effect of IPT-trauma (IPT-T) on PTSD, demonstrated in a randomized effectiveness trial, was mediated by improved social function. Method: The randomized effectiveness trial compared IPT-T to clinic psychotherapy in a publicly funded community mental health center. Women with CSA histories and major depression (n = 162; Mage = 36.3; 54% White, 38% Black, 8% other race) were randomly assigned to IPT-T or CP, and 16 free sessions of IPT-T or CP were offered within a 32-week treatment period. Eighty-eight percent of the sample met diagnostic criteria for PTSD. Results: Mediation models confirmed hypotheses, showing that the effect of IPT-T on improved PTSD symptoms at Week 32 (end of treatment) was partially mediated by better social functioning at Week 16 (middle of treatment period; path c1: B = 12.25, p = .04, 95% confidence interval [.44, 24.05]). Conclusions: Findings support that a non-exposure-based therapy can reduce PTSD symptoms via improved social functioning. Given extensive literature showing the importance of social functioning in mental health generally and PTSD specifically, integrating a focus on social functioning (e.g., building social skills, expanding support networks) into treatments, including exposure-based treatments, may enhance outcomes.