The strategy of the Epstein-Barr virus to persist lifelong in the host depends on establishing a reservoir, which cannot be detected by the immune system but allows reactivation of the virus for shedding and transmission to a new host. Epithelial cells and B-cells play a major role in this viral strategy of EBV, since differentiating epithelial tissues were shown to be permissive for lytic replication in vivo, whereas the B-lymphocytes become predominantly latently infected. However, which cells are the reservoir and which the sites of lytic replication are not quite clear. With the technique of reverse transcription, PCR and immunohistochemistry, we demonstrated that the B-cells of the peripheral blood are a major site of virus production during the primary infection during infectious mononucleosis. These permissive B-cells were also detected after convalescence, however, the absence of any lytic transcripts suggested an efficient immunological control very early in the viral lytic cycle. Serological data on reactivation of EBV correlated with the detection of lytic cycle transcripts in the blood and thus demonstrated that the site of virus production during infectious mononucleosis must be different from that of the persistent state. In those cases, where the infection takes a chronic active course, control of lytic replication is insufficient, either on the level of immune surveillance or of viral gene regulation. We have demonstrated a virus strain with a lytic phenotype in an individual suffering chronic active infection. The impaired capability of this virus to immortalise B-cells correlated with an enhanced expression of the lytic switch gene BZLF-1 and down-regulation of latent regulatory genes in the early phase of infection.