Human immunodeficiency virus (HIV) has been implicated in neurological complications in developed countries. Developing countries have different viral clades and potentially different genetic and social risks for these complications. Baseline neurological performance measures associated with HIV infection have rarely been available from developing countries. The authors carried our a cross-sectional neurological evaluation of a cohort of community-dwelling treatment-naïve HIV-infected patients and similar control subjects from the same communities in Ethiopia. Blinded evaluation using standardized structured questionnaires and a neurological examination was performed by neurologists and treating physicians trained by an HIV neurology specialist. Quantitative performance measures for cognitive and motor function were employed. Data were analyzed with descriptive statistical methods, standard contingency table methods, and nonparametric methods. HIV-positive and control groups were similar by age, gender, and job site. Participants included 73 HIV-positive and 87 HIV-negative controls. Fingertapping speed in the dominant hand was more poorly performed in HIV positives than negatives (P = .01) and was significantly associated with HIV viral load levels (P = .03). Other quantitative neuropsychiatric tests including timed gait, grooved pegboard, task learning, and animal naming did not show significant differences between the two groups. The overall prevalence of central nervous system (CNS) and/or peripheral nervous system (PNS) disease did not significantly differ in the two populations. HIV patients had slowed fingertapping speed correlating with viral load. Other measures of CNS and/or peripheral nervous performance did not differ from controls. The unanticipated minor evidence of HIV-associated neurocognitive and peripheral nerve deficits in this untreated HIV-positive population invite further investigation.