OBJECTIVE: The purpose of this study was to evaluate for APA a proposed strategy to diagnose somatization disorder for possible inclusion in DSM-IV. METHOD: Five sites--Washington University, University of Kansas, University of Iowa, University of Arkansas, and Mount Sinai Medical Center in New York--participated in a collaborative field trial. Female subjects (N = 353) were recruited from several different services (psychiatry, internal medicine, and family practice) and were evaluated for the presence or absence of the disorder. This assessment was performed with a new instrument constructed by combining all the criteria for somatization disorder from the proposed criteria for DSM- IV, DSM-III, DSM-III-R, Perley-Guze, and proposed criteria for ICD-10. RESULTS: A high level of concordance was found between the proposed diagnostic strategy for DSM-IV and the current criteria (DSM-III-R), as well as the earlier criteria (Perley-Guze and DSM-III). The ICD-10 criteria agreed poorly with all other criteria sets. The level of experience of the rater (expert versus novice) with the earlier (Perley- Guze, DSM-III) and current (DSM-III-R) criteria did not influence the identification of cases by use of DSM-IV criteria. No racial effect was introduced by any of the criteria sets. CONCLUSIONS: The strategy for DSM-IV is an accurate and simpler method of diagnosing somatization disorder that does not require special expertise for proper use.