Although some would deny the importance of research in resource-poor countries, the benefits of research to implementation of treatment for HIV infection are innumerable. These benefits include the development of infrastructure, training of staff, creation and validation of algorithms appropriate for the setting, and answering questions necessary for a safe and effective roll-out of therapy. This was true in the USA in 1986, 1 year after the antibody test for HIV was developed, and is true in Africa today. Shortly after the development of the HIV antibody test and before any antiretroviral therapy, few physicians or centres were willing to provide care for HIV patients and fewer had adequate facilities to do so. At that time it was not known how to make an adequate diagnosis of many of the opportunistic infections nor was there a clear idea of how to treat the patients. No-one knew either the best or most cost-effective method to prevent infections. Even as roll-out of therapy proceeded in early 1987 with the approval of zidovudine by the US Food and Drug Administration, physicians were clueless as to when to start treatment. With the addition of other medications in the armamentarium, clinicians began to make mistakes in their ignorance, adding on medications one at a time as they were approved, which led to accumulation of resistance mutations for a generation of patients. These mutations were transmitted to partners and children. What single-handedly helped advance treatment in the USA and Europe in the 1980s was the willingness of respective governing authorities to create clinical research groups not only to develop new drugs but to help create cost-effective ways to use them. All the current treatment guidelines were developed from that research. Over the years these research groups provided care, including medications, laboratory tests and physician and nurse time, for thousands of patients. Medical centres, where these indigent patients were receiving their care, were encouraged to open their doors, creating state of the art clinics and inpatient wards. A generation of clinicians was trained at these research centres where the bulk of US HIV patients were treated. They provided care as they were conducting research. The ability of resource-poor countries to deliver large-scale roll-out plans is dependent on the development of leadership and skills to implement the programmes. South Africa, despite a delay in initiating a national treatment programme, is an example of a country where the research conducted in the period 1996 to 2004 has enabled a skilled set of clinicians, pharmacists and paramedical staff to provide leadership in the scale up of antiretroviral therapy programmes. Guideline development, training and implementation have been led by treatment experts who learned their skills in the research arena.