To evaluate the role of perinatal transmission in the spread of hepatitis C virus (HCV), we screened a cohort of pregnant intravenous drug using (IVDU) women for HCV antibody detection; where seropositive HCV RNA detection by polymerase chain reaction (PCR) was found we followed the infants longitudinally for HCV antibody and HCV RNA. Serum prevalence for HCV for this population was 80% with HCV RNA detected in 50%. Recruitment and follow-up over a 3-year period of a cohort of 83 seropositive women, their 91 newborns and 16 siblings of newborns, showed that there had been a 3% perinatal transmission rate with 1 sibling also infected. These positive cases were defined as transient in 1 case (HCV RNA positive by PCR at 1 month, but seronegative and HCV RNA negative at 10 months of age), 2 unevaluable (HCV RNA positive at 2 months of age, but patients lost to follow-up), and 1 chronic infection in a child at 34 months (positive HCV RNA and seropositive 34-month sibling). Maternal HCV RNA levels for those with infected infants was a mean 40-fold greater than those whose babies were uninfected, although this did not reach statistical significance. Of the remaining infants, the majority (93%) had lost passively acquired maternal antibodies by 9 months of age and all by 12 months. Of 18 women who were HCV seropositive and breast feeding (66% of whom were HCV RNA positive in their sera), none had detectable HCV RNA in breast milk. Hence we conclude that transmission of HCV from mother to infant appears to be of low frequency and positivity appears to correlate with maternal circulating viral load.