Many countries in Africa are planning to provide highly active antiretroviral therapy (HAART) to millions of people with acquired immune deficiency syndrome. This will be a highly complex therapy programme. Physician-based models of care adapted from industrialized countries will not succeed in providing treatment to the majority of those who need it in resource-constrained settings. A high priority is to identify care models for Africa that will increase coverage of HAART safely and effectively: key issues are (i) whether nursing staff or non-clinically qualified staff can take the major role in the treatment programme and reduce the workload of physicians, (ii) whether treatment and monitoring can be delivered through peripheral health centres or through home visits and achieve better adherence and be more cost-effective than delivery at hospitals and (iii) which clinical algorithms used by nursing or non-clinically qualified staff will be effective for screening, diagnosing and managing treatment-related side-effects and medical problems being incurred. Many current ART support programmes are making little or no investment in research, but answering important questions on delivery of HAART will be essential if HAART programmes are to be successful in African nations with a high burden of human immunodeficiency virus infection.