Two hundred fifty patients were entered into a multicenter trial to evaluate the efficacy and toxicity of orally administered acyclovir for treatment of recurrent genital herpes. The study consisted of part A, in which patients entered the study within 48 hours of the onset of lesions, and part B, in which patients self-initiated therapy as soon as possible after the onset of a recurrent episode. In both parts, patients received either acyclovir (200 mg) or placebo, five times daily for five days. In both parts, the duration of virus shedding and the time to crusting and healing of lesions were shorter among acyclovir recipients than among placebo recipients. In part B, fewer acyclovir recipients formed new lesions during the study medication period than did placebo recipients. When parts A and B were compared directly, the duration of virus shedding and the times required for crusting and healing of lesions were significantly shorter among acyclovir recipients in part B than among acyclovir recipients in part A. No significant differences in the duration of itching and pain or in the times of subsequent recurrence were noted between acyclovir and placebo groups in either part A or part B. No significant toxic or adverse reactions were seen in acyclovir recipients. Oral acyclovir shortens the duration of virus shedding and the duration of lesions in patients with recurrent genital herpes. These effects are more pronounced when therapy is self-initiated by patients early in the course of a recurrent episode.