Background <p>Despite Uganda making significant progress in reducing mother-to-child transmission (eMTCT), from 20% in 2000 to 2.8% in 2021, regional differences in MTCT rates exist. The drivers of these differences remain unclear in some nomadic populations of the Karamoja region. This study aimed to determine the maternal and infant predictors of HIV infection among HIV-exposed infants (HEI) under 18 months of age in the Karamoja region, Northeastern Uganda.</p> Methods <p>We conducted a retrospective cross-sectional study across twelve public health facilities in Karamoja, Uganda. Data on mothers and their HEI who had their first HIV nucleic acid amplification test at 4 to 6 weeks of age by either dry blood test or point of care testing were abstracted from the Early Infant Diagnosis (EID) register. The primary outcome was an HEI PCR test result (Positive or Negative). Firth’s penalized likelihood regression was used to determine factors associated with Infant HIV positivity. Data was analyzed using STATA version 15, and a p-value of &lt; 0.05 was considered significant.</p> Results <p>A total of 223 mother-baby pairs were included in the analysis, with 52.9% (n=118) of the HEI being female, and the maternal median age was 27 (interquartile range 23-30). The Infant HIV positivity was 2.2% (n=5) and was significantly lower among mothers with no advanced HIV disease (stage 1 or 2)[adjusted odds ratio (aOR): 0.008, 95% confidence interval (CI): 0.0002- 0.35 P=0.012], Mothers having a suppressed viral load (&lt;1000 copies/ml) [aOR: 0.019, 95% CI: 0.0011- 0.36, P=0.008] and infants who received Nevirapine [aOR: 0.038, 95% CI: 0.0015 - 0.922, P=0.044].</p> Conclusion <p>The results showed that maternal viral suppression, absence of advanced HIV disease, and infant receipt of nevirapine prophylaxis were associated with lower infant HIV positivity. Ensuring mother-baby pairs receive all the eMTCT services can avert new HIV infections among the HEI.</p> Clinical trial <p>Not applicable.</p>