BACKGROUND The prevalence of Human immunodeficiency virus (HIV)-associated dementia (HAD) has continued to rise even as incidence has fallen. Several host genetic variants have been identified that modify risk for HAD. However, the findings have not been replicated consistently and most studies did not consider the multitude of factors that might themselves confer risk for HAD. In the current study, we sought to replicate the findings of previous studies in a neurologically and behaviorally well-characterized cohort. METHODS The sample consisted of 143 HIV+ individuals enrolled in the National NeuroAIDS Tissue Consortium (NNTC). Based on consensus diagnosis, 117 were considered neurologically normal upon study entry, and 26 had HAD. Seven single-nucleotide polymorphisms (SNPs) were genotyped within seven genes (CCL2, CCL3, CCL5, interleukin-1α [IL-1α], IL-10, stromal cell-derived factor 1, and tumor necrosis factor-α). Logistic regression analysis was used to predict group membership (normal vs HAD), with predictor variables including length of infection, age, current drug dependence, current depression, and genotype. RESULTS The two groups were statistically similar with regards to demographic characteristics, current drug use, and disease factors. The HAD group had significantly greater number of individuals with current depression. Only one SNP, rs1130371 within the gene for CCL3, was entered into the analysis as the others showed symmetric distribution between groups. Logistic regression indicated that current depression and CCL3 genotype were significant predictors of HAD. Depression conferred a fivefold greater risk of HAD, while the TT genotype for CCL3 SNP (rs1130371) was associated with twofold risk for HAD. CONCLUSION Depression and CCL3 genotype predicted HAD. The fact that SNPs previously found to be associated with HAD were not in our analysis, and that rs1130371 is in high linkage disequilibrium with neighboring genes indicates that more dense genotyping in significantly larger cohorts is required to further characterize the relationship between genotype and risk for HAD.