Since the introduction of combination antiretroviral therapy, there has been a decrease in both AIDS-defining illnesses and deaths. This decrease meant that performing clinical trials with clinical outcomes in HIV infection became more time consuming and hence costly. Improved understanding and knowledge of HIV led to short-term trials using surrogate outcomes such as viral load and CD4 count. This established a faster drug approval process that complimented the rapid need to evaluate and provide access to drugs based on short-term trials. However, no treatment has yet been found that eradicates the infection, so when treatment is started it is currently a lifelong commitment. Is it reasonable then that guidelines are based almost completely on short-term randomized trials and observational studies of surrogate markers, or is there still a need for trials with clinical outcomes?