Background: Despite high level, persistent exposure to virus, the majority of children born to chronically HIV-infected women escape infection. Most infections that do occur are initiated by a single genetic variant selected from a diverse quasi-species swarm. Given the relative paucity of activated, HIV-susceptible CD4+ target cells, particularly in the fetus, we sought to determine the minimum level of CD4 expression required for cell entry by early transmitted isolates, and whether we could find evidence for selective transmission of variants particularly adept at utilizing low levels of CD4 (such as is present on macrophages) for productive infection. Methodology: Pseudotype viruses were produced in 293T/17 cells using standard techniques. Affinofile cells, a 293 variant with independently inducible CD4 and CCR5 expression systems and an HIV-inducible secreted luciferase reporter, were treated to maximize expression of CCR5 while CD4 was varied across an ~2 log10 gradient. Infection was normalized as a percentage of entry under saturating CD4 conditions. Percent entry was then plotted by absolute number of CD4 molecules per cell to generate an infection curve. Infection curves were used to back-calculate the number of CD4 molecules/cell necessary to achieve 20% of maximal entry (EC20) for each viral variant. EC20s for epidemiologically linked mother-infant transmission pairs were then analyzed using the GEE model. Results: EC20 values were generated for 160 unique envelope clones from 7 in utero and 11 breast milk transmission pairs. Regardless of transmission route, infant variants required more CD4 molecules/cell to achieve entry as measured by EC20 (p<0.0001). There were no differences in EC20 between variants from infants infected in utero or by breastfeeding (p=0.7228), or between variants from the transmitting mothers in each group (p=0.7340). Conclusions: Our data indicate that vertical transmission of HIV does not select for highly efficient usage of CD4, and in fact may require the presence of susceptible target cells with relatively high CD4 expression for either initiation or propagation of infection. Given the lack of route specificity, the described selection is likely not occurring at the site of initial barrier breach (e.g. placenta or gut epithelium), and may represent a common pathway of amplification/outgrowth required for efficient dissemination following the initial seeding event. Our data also support a role for transient inflammation in enhancing fetal/infant susceptibility to infection.