During prospective studies of infants born vaginally to women with cervical Chlamydia trachomatis infection, we evaluated 27 infants given 0.5% erythromycin ointment and 93 given 1% silver nitrate solution as eye prophylaxis, according to the preference of the parents or delivery room personnel. The cumulative proportion of infants developing chlamydial conjunctivitis was 25% for both groups (P = 0.37, Mantel-Cox test). The cumulative proportion of infants developing chlamydial infection at any anatomic site was 74% for those given erythromycin and 70% for those given silver nitrate (P = 0.93). The two groups did not differ significantly in cumulative proportions developing nonchlamydial conjunctivitis. These results indicate that, as it is routinely used in our hospital, erythromycin ointment was not more effective than silver nitrate as prophylaxis against chlamydial conjunctivitis. The influence of delayed administration on efficacy of prophylaxis requires further evaluation.