OBJECTIVE: The aim of this study was to investigate the effect of antiretroviral drugs on neonatal serum insulin levels. RESEARCH DESIGN AND METHODS: A prospective study was conducted on 57 pregnant women divided into three groups: the zidovudine (ZDV) group, HIV-infected women taking ZDV (n = 20); the triple treatment group, HIV-infected women taking triple antiretroviral agents ZDV + lamivudine + nelfinavir (n = 25); and the control group, pregnant women considered normal from a clinical and laboratory standpoint (n = 12). Blood was collected from the umbilical cord of newborn infants upon delivery for measurement of insulin level. The insulin measurements were performed in duplicate by radioimmunoassay. RESULTS: Demographic and anthropometric data were homogeneous, and pregnant women with a personal and family history of diabetes were excluded. There was no difference between groups regarding glycemia in the newborn. Median newborn insulin doses were 2.9, 4.8, and 6.5 muU/ml for the triple treatment, ZDV, and control groups, respectively (P < 0.05). CONCLUSIONS: Use of triple therapy during pregnancy induced a significant decrease in serum levels of neonatal insulin compared with the control group. Active surveillance of short- and long-term adverse events is imperative to issue a definitive statement regarding the impact that use of protease inhibitors during pregnancy will have on infant metabolism.