• Antibiotic usage patterns were studied in two nonproprietary nursing homes that included 720 intermediate care and skilled nursing home beds. Medical records of residents receiving antibiotics were reviewed every fourth month for 1 year. Of 181 antibiotic prescriptions written for indications other than prophylaxis, 41% were for presumed urinary tract infections, 35% for respiratory tract infections, and 14% for skin/soft-tissue infections. The majority of antibiotic prescriptions (54%) were made by telephone order. Cultures were obtained in 60% of suspected infections; two thirds of cultures were of the urine. Antibiotics were changed during the course of therapy in only 12% of cases. Eighty-one percent of residents treated with antibiotics improved or were cured, 9.5% were hospitalized or died, and an additional 9.5% failed to improve but remained in the nursing home. Fever was present in 48% of cases prior to treatment, but had no predictive value for patient outcome. We conclude that antibiotic treatment in the nursing home is often initiated In the absence of fever, culture information, or examination of the patient. Empiric prescription of antibiotics in this setting generally is associated with favorable clinical course. (Arch Intern Med. 1990;150:1465-1468)