For unclear reasons, about 50% of HIV-infected subjects harbour CXCR4-using (X4) viral strains in addition of CCR5-using (R5) viral strains at late stages of the disease. One hypothesis is that a low CD4(+) T-cell surface CCR5 density could facilitate the emergence of X4 strains. Alternatively, one could argue that a high CD4(+) T-cell surface CXCR4 density that is observed in individuals presenting with X4 strains, could favour R5 to X4 switch. Here, we tested both hypotheses. In vivo, we observed by quantitative flow cytometry no difference in CD4(+) T-cell surface CCR5 densities between patients with or without X4 strains. In the course of an in vitro R5 infection, the delay of emergence of X4 mutants was similar between cells expressing 2 distinct cell surface CCR5 densities, but shorter (12 ± 0 days and 21 ± 0 days, respectively, P = 0.01) in cells expressing a high surface CXCR4 density as compared with cells with a low surface CXCR4 density. These data argue for a role of CXCR4 density, but not of CCR5 density, in the emergence of X4 strains. They are reassuring concerning the risk of inducing an R5 to X4 switch using CCR5 antagonists to treat HIV infection.