Measurement of posttraumatic stress disorder (PTSD) in community samples raises particular problems due to the lower base rate of PTSD than in clinical samples. In this paper we describe the development of the DSMPTSD-III-R and the DSMPTSD-IV self-report measures of PTSD for use in community populations. In three groups of motor vehicle accident (MVA) subjects with various rates of PTSD we examine the diagnostic accuracy of the DSM-III-R, DSMPTSD-IV, and three frequently used PTSD measures?Impact of Event Scale (IES), Symptom Checklist-90-Revised (SCL-90-R), and the Minnesota Multiphasic Personality Inventory-Keane PTSD Scale (MMPI-PK)?compared with the Structured Clinical Interview for DSM-III-R (SCID) and its DSM-IV revision. We present two abbreviated versions of our measures, the BPTSD-12-item and BPTSD-6-item, standardized for both DSM-III-R and DSM-IV. Compared with other self-report measures, the DSMPTSD-III-R and DSMPTSD-IV performed well for acute and chronic PTSD in community samples, particularly when the prevalence of PTSD was low. Both the DSMPTSD-III-R and the DSMPTSD-IV are internally consistent and valid measures of PTSD in MVA subjects in community samples. Our measures are utilitarian: the measures are brief, require no training to administer, and have good reliability and validity in community samples. They are among the few measures structured to meet PTSD diagnostic criteria for both the DSM-III-R and DSM-IV. The BPTSD-12 and BPTSD-6 are good, brief PTSD screening measures to identify individuals in disaster communities needing further evaluation.