Objective:To compare the efficacy and tolerability of minocycline versus doxycycline in the treatment of nongonococcal urethritis and mucopurulent cervicitis. Design: Randomized, double-blind trial. Setting: Sexually transmitted disease clinics. Patients: 151 men and 102 women with nongonococcal urethritis, mucopurulent cervicitis or whose sexual partner had either condition or a positive culture for Chlamydia trachomatis. Interventions: Minocycline, 100 mg nightly, or doxycycline, 100 mg twice daily, each administered for 7 days. Measurements: At each visit (days 14 +/- 3, 28 +/- 5, and 49 +/- 7) patients were questioned regarding symptoms, signs, drug compliance, and sexual contact. Cultures for C. trachomatis, Ureaplasma urealyticum, and Mycoplasma hominis were obtained at each visit. Results: 253 patients were enrolled (133, doxycycline; 120, minocycline). Chlamydia trachomatis was initially isolated from 31% of men and 39% of women. Men with a positive smear had a higher symptom/sign score (P < 0.001) and were more likely to have chlamydia (P = 0.004). Positive endocervical smears were not associated with symptoms or signs (P > 0.2) but correlated with isolation of chlamydia (P < 0.001). One hundred sixty-two patients (64%) completed the study. The proportion with urethritis or cervicitis did not differ by treatment group at any follow-up visit (P > 0.08). Unprotected sexual contact did not affect clinical or microbiological cure rates. Adverse effects occurred more frequently in the doxycycline group (men: 43% versus 26%; P = 0.05; women: 62% versus 35%; P = 0.009). Although the proportion with dizziness did not differ by drug administered (P = 0.1), dizziness was reported more often by women (11% versus 3%). Conclusions: Minocycline, 100 mg nightly, was as effective as doxycycline, 100 mg twice daily, each given for 7 days in the treatment of nongonococcal urethritis and mucopurulent cervicitis. Vomiting and gastrointestinal upset occurred more frequently in the doxycycline group.