The failure to see HIV/AIDS in context has been rightly lamented. Although much work has already been done to give us a clearer picture of the complex social correlates of AIDS, there are areas which remain under-researched. A comprehensive assessment of the contextual backdrop to this disease must seek to analyse not only its social determinants but also the effect (or lack of effect) of measures already taken to contain it. The assessment must also pay attention to the social and institutional context within which such measures have been implemented. 1

The failure to consider the effectiveness of existing measures as a relevant part of the context of AIDS is evident not only through its absence from the literature. It is also apparent from discussions with health and education practitioners working in Durban, the site of this study. Many of them feel that a lack of research into the effects of their programmes means they are destined to work with little sense of how, and to what extent, they are succeeding in in uencing AIDS-related behaviour. Although compendia of effective AIDS intervention models do exist and are debated in various fora, 2 on-the-ground practitioners point out that their hectic schedules and the stringent conditions set by individual funders impede broader collaboration, thus limiting the extent to which they can learn from each other’s and their own experiences. While each separate life skills peer education programme has to give an account of its activities to its own specific funders, there is little external or independent assessment of the broader, overall impact of these efforts.