The Ambulatory Sentinel Practice Network (ASPN) conducted an observational study of pelvic inflammatory disease (PID) in the primary care setting. During 14 months from 1982 to 1983, 38 practices in 16 states and two Canadian provinces reported 384 first visits for patients with PID. PID, as diagnosed by the clinicians in ASPN, was less severe than other published clinical descriptions of PID. Fewer patients had fever, palpable adnexal masses, and extensive tenderness than reported in other series. Although 43% of patients met published recommendations for hospitalization, the clinicians hospitalized only 9% on the initial visit, similar to the hospitalization rate by office gynecologists in a national study. Whether this management represented optimal care in the primary care setting is unknown. If office based physicians in the United States admitted to the hospital all PID patients who met current recommendations for hospitalization, the added annual cost could exceed $1.2 billion. Given both the serious clinical consequences of PID and the enormous financial implications of different clinical strategies, there is a compelling need to investigate the diagnosis and management of PID in primary care.