Though antiretroviral therapy attenuates neurocognitive disruption, impairment is still observed. We studied the nadir CD4 cell count as a predictor of neurocognitive changes. This cross-sectional study assessed 64 HIV-infected patients in two groups: G1 (n = 26, nadir CD4 < or =200 cells/ml) and G2 (n = 38, nadir CD4 >200 cells/ml). Percentages of patients showing neurocognitive impairment were compared according to different nadir CD4 cutoffs (200, 250, 300, and 350 cells/ml). From G2, we also took the subgroup of patients receiving treatment (G3) and compared this group with G1, in which all patients were being treated. Demographic and clinical variables were evaluated, as were differences in neurocognitive function. Neurocognitive impairment tended to be more prevalent in G1 [19 patients (73.1%)] than in G2 [20 (52.6%), p = 0.123]. When nadir CD4 cutoffs were compared, there was a trend toward more impaired subjects as the CD4 nadir decreased. Significantly different functioning was found in attention/working memory (digit span backward, p = 0.032) and executive functions (trail making test, part B, p = 0.020), with better performance in G2. Comparison between G1 and G3 confirmed those findings. We found differences in neurocognitive functioning in relation to nadir CD4 count in HIV-infected patients. Attention should be given to this value in the management of neurocognitive protection in HIV infection.