Increasing recognition of the fibromyalgia syndrome together with concerns about limitations of currently available criteria led most centers engaged in fibromyalgia research in Canada and the United States to undertake a multicenter effort to define epidemiologically correct criteria for the diagnosis of fibromyalgia. Five hundred fifty-eight consecutive patients (293 with fibromyalgia and 265 controls) were recruited from 16 private practice and university centers. The study used training sessions to increase interrater reliability, and included methods to determine reliability of examination and historical data. Standardized definition and methods of data acquisition by independent, blinded assessors were employed.