The expanded use of multiple antiretroviral drugs during pregnancy has led to a reduction in the occurrence of perinatal transmission of HIV to <2%, but has led to concerns regarding both short-term toxicity and the long-term impact on the woman and her child. Enhanced toxicity of nevirapine has been noted among women with CD4+ lymphocyte counts >250 cells/microL at treatment initiation and among pregnant women on long-term didanosine and stavudine. These drugs should be avoided in such situations if alternatives are available. Efavirenz has been associated with birth defects in monkeys, and several cases of neural tube defects have been reported in humans after first trimester exposure, so treatment with this drug should be avoided during the first trimester. Protease inhibitors have been associated with an increased risk of maternal glucose intolerance, pre-eclampsia and preterm birth in some, but not all, studies. Pregnancies exposed to antiretroviral therapy should be registered with the Antiretroviral Pregnancy Registry as early in pregnancy as possible in order to provide data on the risk of birth defects after exposure. The pharmacokinetics of nucleoside and non-nucleoside reverse transcriptase inhibitors are not significantly changed in pregnancy, so standard dosing may be used. However, concentrations of several protease inhibitors are lower in pregnancy, so ritonavir-boosting or increased doses are required. Of great theoretical concern is the impact of resistance mutations that develop following single-dose nevirapine therapy on the response to later therapy among women and their infected infants. The use of dual nucleoside therapy for 3-7 days after single-dose nevirapine in the mother reduces but does not eliminate the risk of nevirapine resistance; alternative regimens for prevention of resistance are under study, as are the subsequent responses of the mother and her infant to therapy. Short courses of prophylactic zidovudine and nevirapine have been well tolerated in neonates. Concern has been raised, however, that these exposures may lead to persistent mitochondrial dysfunction or later cancers, underscoring the need for long-term surveillance of antiretroviral-exposed, HIV-uninfected infants.