Background: HIV-uninfected infants born to HIV-infected mothers (HIV-exposed uninfected, HEU) have been described to have immune alterations as compared to unexposed infants. This study sought to characterize T-cell populations after birth in HEU infants and unexposed infants living in a semirural area in southern Mozambique. Methods: Between August 2008 and June 2009 mother-infant pairs were enrolled at the Manhiça District Hospital at delivery into a prospective observational analysis of immunological and health outcomes in HEU infants. Infants were invited to return at one month of age for a clinical examination, HIV DNA-PCR, and immunophenotypic analyses. The primary analysis sought to assess immunological differences between HEU and unexposed groups, whereas the secondary analysis assessed the impact of maternal HIV RNA viral load in the HEU group. Infants who had a positive HIV DNA-PCR test were not included in the analysis. Results: At one month of age, the 74 HEU and the 56 unexposed infants had similar median levels of naïve, memory and activated CD8 and CD4 T-cells. Infant naïve and activated CD8 T-cells were found to be associated with maternal HIV-RNA load at delivery. HEU infants born to women with HIV-RNA loads above 5 log10 copies/mL had lower median levels of naïve CD8 T-cells (p = 0.04), and higher median levels of memory CD8 T-cells, (p = 0.014). Conclusions: This study suggests that exposure to elevated maternal HIV-RNA puts the infant at higher risk of having early T-cell abnormalities. Improving prophylaxis of mother to child HIV programs such that more women have undetectable viral load is crucial to decrease vertical transmission of HIV, but may also be important to reduce the consequences of HIV virus exposure in HEU infants.