The importance of the vpr gene for simian immunodeficiency virus (SIV) replication, persistence, and disease progression was examined by using the infectious pathogenic molecular clone called SIVmac239. The ATG start codon of the vpr gene was converted to TTG by site-specific mutagenesis. The constructed Vpr- mutant virus is identical with the parental SIVmac239/nef-stop virus with the exception of this one nucleotide. These viruses replicated with similar kinetics and to similar extents in rhesus monkey lymphocyte cultures and in the human CEMX174 cell line. Five rhesus monkeys were inoculated with the Vpr- variant of SIVmac239/nef-stop, and two monkeys received SIVmac239/nef-stop as controls. Both controls showed reversion of the TAA stop signal in nef by 2 weeks postinfection, as has been observed previously. Reversion of the TAA stop codon in nef also occurred in the five monkeys that received the Vpr- variant, but reversion was delayed on average to about 4 weeks. Thus, the mutation in vpr appeared to delay the rapidity with which reversion occurred in the nef gene. Reversion of the TTG sequence in vpr to ATG was observed in three of the five test animals. Reversion in vpr was first observed in these three animals 4 to 8 weeks postinfection. No vpr revertants were found over the entire 66 weeks of observation in the other two test animals that received the vpr mutant. Antibodies to vpr developed in those three animals in which reversion of vpr was documented, but antibodies to vpr were not observed in the two animals in which reversion of vpr was not detected. Antibody responses to gag and to whole virus antigens were of similar strength in all seven animals. Both control animals and two of the test animals in which vpr reverted maintained high virus loads and developed progressive disease. Low virus burden and no disease have been observed in the two animals in which vpr did not revert and in the one animal in which vpr reversion was first detected only at 8 weeks. The reversion of vpr in three of the five test animals indicates that there is significant selective pressure for functional forms of vpr in vivo. Furthermore, the results suggest that both vpr and nef are important for maximal SIV replication and persistence in vivo and for disease progression.