PIP: Nucleotide sequences were obtained for portions of the envelope and gag genes of 4 HIV-1 isolates from Nigeria. The gag gene sequences clustered with previously described gag sequences from Gabon, Zaire, and Taiwan, which form the G clade of HIV-1. This documented for the first time the presence of subgroup G viruses in Nigeria. The env gene sequences were most closely related to env sequences reported from 2 isolates previously classified as gag subgroup G and, with those 2 env sequences, defined the subgroup G env genotype. Virus isolates were obtained by cocultivation of peripheral blood mononuclear cells (PBMCs) from Nigerian AIDS patients (isolates JV1018 and JP88) or healthy prostitutes (G3 and G9) with normal uninfected PBMCs (JV1083, JP88, and G3) or GEM-SS, a T cell line (G9). Clones containing env and gag gene sequences were obtained by polymerase chain reaction amplification, using DNA from these cultures. The region of gag amplified by primers SK37 and SK39 containing BamHI and KpnI sites, respectively, in 5' tails, was cloned into the KpnI-BamHI site of pKSBluescript and sequenced. When the gag sequences from 2 of the Nigerian viruses were subtyped by weighted parsimony, both viruses grouped with others of the gag G subgroup. A similar analysis of env sequences from all 4 Nigerian isolates showed that they were most closely related to 2 env sequences that had been assigned to env subgroup G. The Nigerian env sequences clustered more closely with each other than they did with the other 2 subgroup G viruses. Consensus of the inferred amino acid sequences of the 4 env genes showed that the V3 loop closely resembled the A to F consensus. 8 positions in the Nigerian consensus sequence appeared to be unique. However, whether these are signature residues for Nigerian subgroup G isolates or for subgroup G isolates will require the characterization of more isolates.