Despite the preferential localization of CD4 to lipid rafts, the significance and role of these microdomains in HIV-1 entry is still controversial. The possibility that CD4, when localized to non-raft domains, might be able to support virus entry cannot be excluded. Because disintegration of rafts by extraction of cellular cholesterol with methyl--cyclodextrin suffers from various adverse effects, we investigated molecular determinants controlling raft localization of the CD4 receptor. Extensive mutagenesis of the receptor showed that a raftlocalizing marker, consisting of a short sequence of positively charged amino acid residues, RHRRR, was present in the membrane-proximal cytoplasmic domain of CD4. Substitution of the RHRRR sequence with alanine residues abolished raft localization of the CD4 mutant, RA5, as determined biochemically using solubilization in nonionic detergents and by confocal microscopy. The possible inhibitory effect of the introduced mutations on the adjacent CVRC palmitoylation site was ruled out because wild type (wt) CD4 and RA5, but not a palmitoylation-deficient mutant, were efficiently palmitoylated. Nonetheless, the RA5 mutant supported productive virus entry to levels equivalent to that of wild type (wt) CD4. Sucrose gradient analysis of Triton X-100 virus lysates showed that Gag and envelope gp120 proteins accumulated in low buoyant, high-density fractions. This pattern was changed after virus incubation with cells. Whereas Gag proteins localized to lipid rafts in cells expressing wt CD4 and RA5, gp120 accumulated in rafts in cells expressing wt CD4 but not RA5. We propose that raft localization of CD4 is not required for virus entry, however, post-binding fusion/entry steps may require lipid raft assembly.