The challenges facing the health system in South Africa are likely to impact on life-long adherence for patients in the context of the rollout of ART. Smaller ART programs have been able to demonstrate good adherence rates, but the question remains if this can be achieved by large public sector ART programs. Most adherence researchers share the basic understanding that patients are adherent when they take their medications as prescribed by the health provider. An approach to adherence that combines both clinical and social knowledge—a biosocial approach— is likely to move us to a better understanding of adherence and how to improve adherence to ART. This study on social support and ART adherence aims to gather and document information that could be used to improve services and program strategies for strengthening and maintaining adherence at ART rollout sites in Gauteng. The two study sites Carletonville Hospital and Zola Clinic were chosen randomly from all second-generation rollout sites in the Province. Data were collected from a total of 359 respondents, 164 in Carletonville and 195 in Zola. The response rate was 98.3%. The results showed that the majority of the respondents were female (72.1%) and about 44.9% were within the age group 30-39 years. In terms of educational attainment, most respondents (70.1%) had received secondary education and 2.5% had not attended school. Based on assets quintiles scores of 1-5, with 5 being the highest score, about one-third of the respondents scored 1, and only 7% scored 5. Compared with Carletonville, respondents from Zola were more educated and better resourced. At the facilities, treatment preparation and support and adherence assessment procedures are routine features of the ART program and entail pre and post test counseling, group education and adherence counseling and serve as mechanisms for adherence support. This is enhanced by routine follow-up appointments where ART patients are provided information on side effects of ARVs, effectiveness of treatment, CD4 cell and viral load counts and referral to services not provided at the facility. Additionally, support groups accessed by patients undertake a range of educational activities on staying healthy, viral load and CD4 cell counts and ARVs. Although respondents were largely positive about their interactions with health providers and the support they provided, some expressed concern about health workers being too busy to address their problems, not treating patients with enough respect and sometimes patients leaving without receiving treatment because staff were either absent or late or queues were too long. The HIV disclosure rate was high (95.5%). However, respondents were more likely to disclosure to a family member, but less likely to a friend, neighbor or religious leader. Selfreported adherence and viral load adherence rates were high (97.6% and 76.6% respectively) but CD4 adherence was lower at 51.0%. The study did not document a convincing association between social support and ART adherence. Only two variables (receiving food supplements and age groups) were significantly associated with CD4 and viral load adherence. Given the limitations of the study, a longitudinal study is needed in these sites to better understand the predictors of short and long-term adherence and to explore ways to better measure the relevance, content and quality of the social support services being utilized by ART patients at facility and community levels. Interventions and policies are needed to respond to the concerns identified from the study regarding inadequate attention and respect by health providers, absence or lateness of doctors and pharmacists and challenges pertaining to access to food, income and disability grants.