Objectives The study aims to describe the socio-demographic characteristics, clinical outcomes of the patients in the Gauteng public sector roll-out programme and establish the association between these. There are contradictory results from international studies on these associations, in the absence of SA results. Methods This is a retrospective cohort, exploratory, secondary data, record review study and a comparison between two sites. Routinely collected socio-demographic data and clinical data were used to establish the impact of socio-demographic factors on response to HAART. This was collected for patients who enrolled from April 2004 to August 2004. Chris Hani Baragwanaath (CHB) had 494 records, Helen Joseph (HJ) had159 records collected. Exposure variables (age, sex, marital status, education level, residential area, employment, baseline viral load and baseline cd4 count). Outcome variables were (CD4 and Viral load at 3 months, 6 months and 12 months). Data Analysis T tests were used for comparing means; logistic regression was used to find the effect of ordered exposure variables and binary outcome. Chi square and fishers exact were used to find frequencies and association between the categorical variables. Regression was used to find the association between the continuous exposure variables and the continuous outcome variables. In a multivariate model, to assess the effect of the exposure variables to the outcome variables Multivariate regression was used. Statistical significance was assessed at the 5% significance level, giving 95% confidence interval. Results The majority of the patients (653) were female, African, unemployed and were literate. At CHB, at the end of the first year, three quarters were still on treatment however; just under a fifth (19%) had died. The majority responded well to treatment and had a mean baseline CD4 count of 58.9cells/mm3 (CHB) and 78.4cells/mm3 (HJ) and mean CD4 count of 245 (CHB) and 268 (HJ) after 12 months. increasing age, and being widowed, lowers the immunological response. Employment, education, sex and had no impact on response. Conclusion • There is positive virological and immunological response to HAART in Gauteng ARV roll-out programme despite the low socio economic status of the majority of the patients. • Provision of free antiretroviral drugs and access to the disability grant has assisted in mitigating the effects of HIV/ AIDS on the socio-economically disadvantaged. • The elderly and the widowed might need close monitoring as their response appears to be lower than the others. • The group with no schooling is not well represented in this sample; the question is whether the HIV/AIDS prevention messages and treatment is accessible for this group. This needs further research.