Public debates on HIV/AIDS in South Africa have, for the last three years, been dominated by the controversy around the causal link between HIV and AIDS. A decision on this question has direct consequences for health policy, treatment, and education. However, this question also warrants a medical-historical investigation into concepts and models of causality and the way they have panned out in medical-scientific revolutions, and in diagnostics and treatment. As this paper attempts to show, necessary and sufficient criteria for disease causation are crucial in the debates on the aetiology of HIV/AIDS. In the course of the history of medical diagnostics, the sufficiency criterion has been considerably modified, while the necessity criterion has been foregrounded. It has been shown that the difficulties surrounding the establishment of strict sufficiency criteria do not preclude the elaboration of an aetiology of HIV/AIDS. While mainstream medical science privileges the necessity criterion, the AIDS dissenters insist on strict sufficiency for conclusive proof of the causal link between HIV and AIDS. This paper aims to show that both criteria have a role to play, but in differentiated ways and at different and distinct sites of intervention.