The diagnostic accuracy of screening instruments in the immediate aftermath of disaster is unknown. Following the 2005 London bombings a 2-year outreach program identified and screened affected persons. In 331 persons the accuracy of the Trauma Screening Questionnaire (TSQ) was compared to a DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; American Psychiatric Association, 1994) diagnosis of posttraumatic stress disorder. The sensitivity of the TSQ was comparable with previous studies, but specificity was substantially lower. Specificity was lowest immediately after the bombings and rose steadily over the next 18 months. Specificity was also lower in non-White individuals, who had elevated scores on the TSQ. The data are consistent with previous research showing elevated distress in ethnic minorities following trauma, and suggest limitations on the performance of screening instruments.