Background: Drug resistance mutations (DRM) may circulate in HIV RNA and persist in CD4 cell DNA despite antiretroviral therapy (ART). To determine risks for transmission and community reservoirs of DRM, DNA genotype was assessed from PBMC DNA of viremic and suppressed ART recipients in a community clinic population. Methodology: HIV-1 sequences from consecutive consenting patients on ART were enrolled and followed 6 monthly for 2 years. Viremic subjects (>50 copies/ml of RNA) had genotypes from plasma RNA and PBMC DNA, and those suppressed (< 50 copies/ml of RNA) from PBMC DNA only. Phylogenetic relationships were inferred using the Neighbor Joining method, with 2 Kimura parameters and 1000 bootstrap replicates. Statistical comparisons of median values were performed by non-parametric Rank-sum tests. Results: Of the 124 individuals (27 female, 96 male and 1 transgender) 35 were African-American (AA), 30 Hispanic, 12 Asian and 47 White. The 38 viremic subjects had a median viral load and CD4 cell count of 5,865 copies/ml and 298 cells/mm3; 24/38 (63%) were wild-type (WT) and 12 (27%) had the same DRM in both RNA and DNA (concordance of 89%). Two had M184V in DNA but not in RNA and 2 in RNA but not DNA. Among the 12 with DRM, the median CD4 was significantly lower compared to those who were WT, 276 vs 381 cells/mm3 (p<0.01). Among 86 suppressed individuals, 37/86 (43%) had one or more archival DRM; 9 had 2 classes and 1 had 3 classes of DRM. CD4 cells counts were similar between 35 with and 51 without DNA DRM (413 vs 427 cells/mm3). Eleven of 86 (13%) of those suppressed at entry developed virologic failure by 48 weeks. DRM developed in plasma RNA of 6/11 (55%), 3 who were DNA WT and 3 with DNA DRM at entry. Phylogenetic analysis (Fig 1) showed related sequences included 2 pairs of men with bootstrap values of 97 and 99 % and one trio (2 men and 1 woman) 61%. Seventeen older AA injection drug users (IDU) clustered with a low bootstrap value (45%). Conclusions: Analysis of DNA genotype may guide treatment and identify transmission networks. DRM in archival DNA may pose a risk of transmission and re-emergence of resistance at failure and may limit future treatment options. The weak bootstrap linkage among a cluster of 17 older AA men and women who were former IDU, suggests a distant founder effect in this population.