To determine if immunity to human cytomegalovirus (HCMV) protects women from acquiring HCMV from their children, a blinded, randomized protocol was used to monitor seronegative women who received placebo or Towne vaccine (approximately 500 pfu) and seropositive women. Each group was similar for mean maternal (33 years) and child age (18 months) and duration of viral shedding by the child (15 months). Among 19 placebo recipients, 9 developed primary infection; 8 of 19 vaccines but only 3 of 42 naturally seropositive subjects had evidence of acquiring HCMV from their child. Wild type infection and Towne vaccine induced similar mean lymphoproliferative responses to HCMV antigens, but one dose of Towne vaccine produced mean neutralizing titers 10- to 20-fold lower than those after wild type infection. Thus, a vaccine that induces immune responses equal to those induced by wild type virus may protect healthy women from acquiring HCMV from their children.