To estimate Ne for HIV-1 in vivo, several studies have employed idealized models that assume HIV-1 evolution to be neutral that genomic variations do not lead to variation in fitness and therefore selective forces are inconsequential. By comparisons of model predictions with data on polymorphisms in the env or the gag-pol region of HIV-1, the latter studies obtained Ne,102�C104. These latter studies employed several tests to ascertain the predominant neutrality of HIV-1 evolution. More recent evidence, however, points to significant selective pressures on both the env and the gag-pol regions, rendering uncertain the estimates of Ne obtained by neutral models. Rouzine and Coffin considered HIV-1 evolution with selection and predicted the frequency of the least abundant haplotype in a two-locus/two-allele model. By comparison with data from env and pro regions, the latter model yielded Ne.105. The latter model, however, did not include recombination. Growing evidence, including the observation of circulating recombinant forms of HIV-1 as well as recombinant forms unique to individuals, points to the significance of recombination in the evolution of HIV-1. Recombination alters the association of mutations and influences the ICI 182780 prevalence of haplotypes, which in turn may affect the estimate of Ne obtained by Rouzine and Coffin. It is of importance therefore to estimate Ne using a model of HIV-1 evolution that incorporates both selection and recombination. SCH772984 Substantial efforts are ongoing to describe HIV-1 evolution in the presence of recombination. Recent advances in mathematical modelling and stochastic simulations have provided valuable insights into the role of recombination in the genomic diversification of HIV-1 in vivo, particularly in the context of the development of resistance to antiretroviral therapy. Specifically, the influence of recombination is predicted to depend sensitively on Ne and on the nature of fitness interactions between loci, characterized by epistasis: When Ne is small, recombination tends to lower viral genomic diversity independent of epistasis, whereas when Ne is large, recombination lowers diversity if epistasis is positive. Further, recombination is also predicted to lower the waiting time for the emergence of viral genomes carrying new, potentially favourable combinations of mutations. Our aim is to employ a model of HIV-1 evolution that accurately mimics viral genomic diversification in infected individuals as a function of the population size and estimate Ne from comparisons of model predictions with patient data.