The purpose was to estimate the frequency, characteristics, and risk factors of HIV-associated distal sensory polyneuropathy (DSP) among South Africans who attend an urban community-based clinic. In a cross-sectional study, neuropathy status was determined in 598 HIV-infected adults using validated tools (Brief Peripheral Neuropathy Screen and a modified version of the Total Neuropathy Score) to categorize subjects as DSP versus no DSP. Symptomatic DSP (SDSP) required the presence of at least two neuropathic signs together with symptoms. Clinical, anthropometric, and laboratory evaluations were prospectively performed. CD4 counts, antiretroviral therapy (ART), and questionnaires regarding previous tuberculosis (TB) and alcohol exposure were collected retrospectively. Approximately half (49%) of the study population were diagnosed with DSP, and 30% of the study population were diagnosed with SDSP. In multivariate analyses the odds ratio (OR) (95% confidence interval) of DSP were independently associated with ART use (OR 1.7, 1.0-2.9), age (per 10 year increment) (OR 1.7, 1.4-2.2), and prior TB (OR 2.0, 1.3-3.0). Pain or paresthesias were reported as moderately severe by 70% of those with SDSP. Stavudine use was significantly associated with DSP. DSP is a clinically significant problem in urban HIV-infected Africans. Our findings raise the possibility that the incidence of DSP may be reduced with avoidance of stavudine-containing regimens in older subjects, especially with a history of prior TB infection.