The Epstein-Barr virus is etiologically linked with the development of benign and malignant diseases, characterized by their diversity and a heterogeneous geographic distribution across the world. The virus possesses a 170-kb-large genome that encodes for multiple proteins and non-coding RNAs. Early on there have been numerous attempts to link particular diseases with particular EBV strains, or at least with viral genetic polymorphisms. This has given rise to a wealth of information whose value has been difficult to evaluate for at least four reasons. First, most studies have looked only at one particular gene and missed the global picture. Second, they usually have not studied sufficient numbers of diseased and control cases to reach robust statistical significance. Third, the functional significance of most polymorphisms has remained unclear, although there are exceptions such as the 30-bp deletion in LMP1. Fourth, different biological properties of the virus do not necessarily equate with a different pathogenicity. This was best illustrated by the type 1 and type 2 viruses that markedly differ in terms of their transformation abilities, yet do not seem to cause different diseases. Reciprocally, environmental and genetic factors in the host are likely to influence the outcome of infections with the same virus type. However, with recent developments in recombinant virus technology and in the availability of high throughput sequencing, the tide is now turning. The availability of 23 complete or nearly complete genomes has led to the recognition of viral subtypes, some of which possess nearly identical genotypes. Furthermore, there is growing evidence that some genetic polymorphisms among EBV strains markedly influence the biological and clinical behavior of the virus. Some virus strains are endowed with biological properties that explain crucial clinical features of patients with EBV-associated diseases. Although we now have a better overview of the genetic diversity within EBV genomes, it has also become clear that defining phenotypic traits evinced by cells infected by different viruses usually result from the combination of multiple polymorphisms that will be difficult to identify in their entirety. However, the steadily increasing number of sequenced EBV genomes and cloned EBV BACS from diseased and healthy patients will facilitate the identification of the key polymorphisms that condition the biological and clinical behavior of the viruses. This will allow the development of preventative and therapeutic approaches against highly pathogenic viral strains.