OBJECTIVE This report describes medications provided or prescribed during ambulatory medical care visits in 1992. Total ambulatory care medication therapy combines data from office-based physicians, hospital outpatient departments (OPD's), and hospital emergency departments (ED's). Drug therapy is described along three dimensions: number of drugs provided or prescribed (drug mention), whether a visit had any drugs mentioned (drug visit), and average number of drugs mentioned per 100 visits (drug mention rate). Utilization in ambulatory care settings is compared in terms of patient, drug, provider, and visit characteristics. METHODS Annual use of medication therapy was determined using data collected in the 1992 National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS). NAMCS includes office visits to nonfederal physicians principally engaged in office practice. The target universe of NHAMCS includes visits to ED's and OPD's of non-Federal, short-stay, or general hospitals. Sample data were weighted to produce annual estimates. Drug mentions are defined as the number of drugs mentioned on the patient record from. RESULTS An estimated 1.1 billion medications were provided or prescribed at ambulatory care visits in 1992. The setting with the greatest percent of visits with medication therapy was the ED; OPD's had the lowest percent with medications. Patients at the ED were provided more pain relief type drugs. The rate of drug mentions and percent of visits with medications were significantly higher in OPD clinics of general medicine and pediatrics compared with other types of OPD clinics. In office-based settings, physicians specializing in cardiovascular diseases were most likely to prescribe medications. Also, cardiovascular-renal type drugs accounted for the largest percent of office-based drug mentions. Visits with illness diagnoses are most likely to receive medication therapy. Trend data comparing 1980 to 1992 office-based mentions showed significant changes on several characteristics: single-ingredient drug status, physician specialty, and patient age. CONCLUSIONS The profile of patients using office- and hospital-based ambulatory care settings are quite different as is the case-mix of conditions. These differences play an important role in medications utilized. The aging of the U.S. population from 1980 to 1992 appeared to have significant effects on several drug mention characteristics.