Burkitt's lymphoma (BL) was first described as a clinical entity in children in Central Africa by Denis Burkitt in 1958. The particular epidemiological features of this tumor initiated the search for a virus as the causative agent and led to the discovery of EpsteinBarr virus (EBV) by Epstein and coworkers in 1964. It became apparent in the seventies and eighties that the tumor is not restricted to Central Africa, but occurs with lesser incidence all over the world (sporadic BL) and is also particularly frequent in HIV infected individuals, and that not all BL cases are associated with EBV: about 95% of the cases in Central Africa, 40 to 50% of the cases in HIV-infected individuals and 10 to 20% of the sporadic cases harbour the viral information and express at least one viral antigen (EBNA1) and a number of non-coding viral RNAs. In contrast, all BL cases regardless of their geographical origin exhibit one of three c-myc/Ig chromosomal translocations leading to the activation of the c-myc gene as a crucial event in the development of this disease. Although epidemiological evidence clearly points to a role of the virus in the African cases, the role of EBV in the pathogenesis of BL has remained largely elusive. This review summarizes current concepts and ideas how EBV might contribute to the development of BL in the light of the progress made in the last decade and discusses the problems of the experimental systems available to test such hypotheses.