A prospective, randomized, controlled, non-blind clinical trial was conducted to compare the efficacy of monotherapy with ciprofloxacin with that of a combination of clindamycin plus gentamicin in the treatment of patients with acute pelvic inflammatory disease. Pretreatment and post-treatment cervical culture specimens were obtained for Neisseria gonorrhoeae, Chlamydia trachomatis, Mycoplasma hominis, and Ureaplasma urealyticum. Pretreatment and post-treatment endometrial culture specimens were obtained for those organisms plus facultative and anaerobic bacteria. Minimal inhibitory concentrations were determined on all isolates by agar dilution. Clinical resolution of infection was seen in 31 of 33 (94 percent) ciprofloxacin-treated patients compared with 34 of 35 (97 percent) clindamycin/gentamicin-treated patients. N. gonorrhoeae was eradicated in all cases and C. trachomatis in 12 of 13 cases (92 percent). Ciprofloxacin appeared less effective than clindamycin/gentamicin in eradicating bacterial-vaginosis-associated organisms from post-treatment culture specimens obtained from the endometrium. Comparable clinical response was seen with both regimens. The significance of persistent bacterial-vaginosis-associated organisms following ciprofloxacin therapy is unclear. However, since one goal of treatment of pelvic inflammatory disease should be to eliminate organisms from the upper genital tract, ciprofloxacin may not provide optimal single-agent therapy for pelvic inflammatory disease.