In this prospective trial, 130 hospitalized patients with acute pelvic inflammatory disease based on clinical criteria were randomly treated with intravenous gentamicin plus clindamycin (N = 63) or cefoxitin plus doxycycline (N = 67) for at least 4 days, followed by oral clindamycin or doxycycline, respectively, for a total of 14 days. Pre-treatment cultures were obtained for endocervical Neisseria gonorrhoeae and Chlamydia trachomatis, and for endometrial C trachomatis and aerobic and anaerobic bacteria. Overall, 46 subjects (35%) had endocervical cultures positive for N gonorrhoeae. Endocervical and endometrial cultures were positive for C trachomatis in 16 and 6%, respectively. Ninety-five percent of patients had at least one aerobic bacterium, 38% had at least one anaerobic bacterium, and only 2% had no organisms isolated from their endometrium. Fifty-seven subjects taking gentamicin-clindamycin (90.5%) and 64 subjects taking cefoxitin-doxycycline (95.5%) were clinically cured, a nonsignificant difference. Three subjects treated with gentamicin-clindamycin and one treated with cefoxitin-doxycycline required hysterectomy or salpingectomy for cure. Follow-up examinations and cultures were performed in 84% of the subjects. Post-treatment cultures for N gonorrhoeae were negative in all cases tested. Post-treatment endocervical and endometrial C trachomatis cultures were negative in ten of 11 subjects treated with gentamicin-clindamycin and in eight of nine treated with cefoxitin-doxycycline, a nonsignificant difference. We conclude that gentamicin-clindamycin and cefoxitin-doxycycline have similar clinical cure rates for acute pelvic inflammatory disease and are equivalent in eradicating genital N gonorrhoeae and C trachomatis.