OBJECTIVE: To summarize evidence for the prevalence of posttraumatic stress disorder (PTSD) among persons with chronic pain (CP). METHODS: We searched databases for studies published between January 1995 and December 2016, reporting the prevalence of PTSD in persons with CP. Two reviewers independently extracted data and assessed the risk of bias. We calculated the pooled prevalence using a random-effects model and performed subgroup analyses according to pain location, the population and assessment method. RESULTS: Twenty-one studies were included and the PTSD prevalence varied from 0-57%, with a pooled mean prevalence of 9.7%, 95% CI (5.2-17.1). In subgroup analysis, the PTSD prevalence was 20.5%, 95% CI (9.5-39.0) among persons with chronic widespread pain, 11.2%, 95% CI (5.7-22.8) among persons with headache, and 0.3%, 95% CI (0.0-2.4) among persons with back pain. The prevalence in clinical populations was 11.7%, 95% CI (6.0-21.5) and in non-clinical populations 5.1%, 95% CI (0.01-17.2). In studies of self-reported PTSD symptoms, PTSD prevalence was 20.4%, 95% CI (10.6-35.5), and in studies where structured clinical interviews had been used to assess PTSD its prevalence was 4.5%, 95% (CI 2.1-9.3). The risk of bias was medium for most studies and the heterogeneity was high ( = 98.6). CONCLUSION: PTSD is overall more prevalent in clinical cohorts of persons with CP and particularly in those with widespread pain, but may not always be more prevalent in non-clinical samples of persons with CP, compared to the general population. There is a large heterogeneity in prevalence across studies. Future research should identify sources of heterogeneity and the mechanisms underlying the comorbidity of the two conditions.