This study assessed the relationship between changes in CD4 T-cell count, HIV Viral Load (VL) level and evolutionary patterns in Antiretroviral (ARV) resistance mutations of circulating HIV in those with persistent viraemia despite treatment. The study examined the dynamics of change in resistance mutations to explore potential predictors of evolution in the circulating virus over a prolonged period. Four longitudinal blood samples from 29 HIV-infected Australian patients at RPAH are analysed for the presence of genotypically resistant HIV virus. The subjects had CD4 cells counts (40-670 cells/mm(3)), VL levels (1.7-5.7 log(10) copies/ml), drug resistance mutations, and had been treated with ARV regimens for varying periods between February 1998 and June 2005 (i.e., 88 months). Parametric and nonparametric tests were used to examine changes in CD4 T-cell counts and VL levels, and in frequencies of mutations at different sample points. Nonparametric LOESS fitting curves were used to analyse changes in CD4 T-cell counts and VL levels. During the study period, changes in mean values of CD4 cell counts and VL levels in patients displaying an evolving genotypic resistance profile differed from those without evidence of an evolving mutational pattern; applied for both Reverse-Transcriptase (RT) and Protease (PR) gene profiles. In patients with Evolution-of-Resistance (EoR) at the RT site, there was a significant decrease in mean CD4 cell count during the first 36 months (-194 cells/mm(3), P=0.0466) but no significant change in the last 52 months or thereafter (-10 cells/mm(3), P=0.8464). Conversely, in patients with EoR at the PR site, there was no significant change in mean CD4 cell count during the first 36 months (-30 cells/mm(3), P=0.6187) but a significant decrease (-155 cells/mm(3), P=0.0348) thereafter. In patients without EoR at the RT site, there was no significant change in mean CD4 count during the first 36 months (-12 cells/mm(3), P=0.8371), however, a significant increase in mean CD4 cell count occurred thereafter (+242 cells/mm(3), P=0.0077). Similarly, in patients without EoR at the PR site there was no significant change in mean CD4 count during the first 36 months (-19 cells/mm(3), P=0.6647) and there was a significant increase in the last 52 months (+145 cells/mm(3), P=0.0348). The only significant decrease in mean value of VL levels occurred in patients without EoR associated with NRTIs thereafter (-1.33 log(10) copies/ml, P=0.0477), while no significant changes in mean value of VL levels in patients with/without EoR associated with any ARV drugs at any other time points (-0.65 to +0.15 log(10) copies/ml, P=0.1855 to 0.7958). Low CD4 cell counts (<250 cells/mm(3)) and high VL levels (>4.50 log(10) copies/ml) in the early stage of HIV infection, a significant decrease in CD4 cell counts during the first 36 months (-50 or more cells/mm(3)), and high frequencies of mutations during the first 36 months of antiretroviral regimen (>3 mutations) emerged as potential predictive factors of EoR associated with NRTI/PRI agents thereafter. Stable VL in the first 36 months correlated with lack of lability of resistance mutations thereafter.