The effectiveness of azithromycin and doxycycline in the treatment of erythema migrans was compared in a prospective randomized trial. One hundred seven adult patients with typical erythema migrans, examined in the Lyme Borreliosis Outpatients' Clinic, University Department of Infectious Diseases in Ljubljana, were included in the study. Fifty-five patients received azithromycin (500 mg twice daily for the first day, followed by 500 mg once daily for four days) and 52 patients received doxycycline (100 mg twice daily for 14 days). The mean duration of skin lesions after the beginning of treatment was 7.5 +/- 5.9 days (median value 5, range 2-28 days) in the azithromycin group and 11.4 +/- 7.8 days (median value 9, range 2 days--8 weeks) in the doxycycline group (p < 0.05). Borrelia burgdorferi was isolated from erythema migrans in 28 patients before therapy: in 13 out of 52 in the doxycycline group and in 15 out of 55 in the azithromycin group. Three months after therapy, the culture was positive in four out of 13 patients treated with doxycycline and in one of the 15 patients who received azithromycin. A biopsy was repeated in all the patients with a positive isolation from the first skin specimen. During the first 12 months' follow-up, three patients treated with doxycycline but none in the azithromycin group developed major manifestations of Lyme borreliosis, while 15 doxycycline recipients and 10 azithromycin recipients developed minor consecutive manifestations.