We compared the efficacy of two treatment regimens for asymptomatic congenital syphilis. Between June 1989 and July 1991, we prospectively and randomly assigned 169 patients to receive either one dose of benzathine penicillin G or procaine penicillin G for 10 days. There were no significant differences between the treatment groups in regard to birth weight, sex, race, gestational age, Apgar scores, infant or maternal rapid plasma reagin (RPR) titers, fluorescent treponemal IgM antibody, or maternal treatment. Patients were examined at 2 to 3, 6, and 12 months after treatment; treatment failure was defined as clinical signs or persistent laboratory evidence of congenital syphilis. Nine patients were removed from the study during the neonatal period, eight were lost to follow-up, and 152 patients were examined 2 to 3 months or more after treatment. Among these 152 patients, none had clinical evidence of congenital syphilis at follow-up, and all the patients tested at 2 to 3 months after treatment (68 in the benzathine penicillin G group and 61 in the procaine penicillin G group) had at least a fourfold decrease in RPR titers. The RPR became nonreactive in all but three of the infants (two in the procaine penicillin G group and one in the benzathine penicillin G group; all three were 2 to 3 months of age when last tested). We conclude that treatment failure did not occur with either regimen and that there was no significant difference in outcome between the two groups.