Background: A validated screening instrument for PTSD in community dwelling older people is lacking. This study evaluates a newly developed measure, the self-rating inventory for posttraumatic stress disorder (SRIP) on its usefulness in survey research. The predictive value of the SRIP in a community setting is investigated. Methods: In a two-phase epidemiologic design the criterion validity of the SRIP was tested against diagnosis made with the comprehensive international diagnostic interview (CIDI) in 1721 older (55-90 years) inhabitants of the Netherlands. Optimal sensitivity and specificity was determined using a weighted receiver operator characteristic (ROC)-curve. Results: Optimal sensitivity (74.2%) and specificity (81.4%) was reached with a cut-off of 39 points. Limitations: According to a strictly applied CIDI algorithm the number of 'true' cases was limited. Conclusion: Overall findings indicate that posttraumatic stress disorder can be identified adequately in a community-based population of older adults using the SRIP. Use of the SRIP may improve recognition and diagnosis of posttraumatic stress disorder in the community.