Epstein-Barr virus (EBV) is characterized by a bipartite life cycle in which latent and lytic stages are alternated. Latency is compatible with long-lasting persistency within the infected host, while lytic expression, preferentially found in oropharyngeal epithelial tissue, is thought to favor host-to-host viral dissemination. The clinical importance of EBV relates to its association with cancer, which we think is mainly a consequence of the latency/persistency mechanisms. However, studies in murine models of tumorigenesis/lymphomagenesis indicate that the lytic cycle also contributes to cancer formation. Indeed, EBV lytic expression is often observed in established cell lines and tumor biopsies. Within the lytic cycle EBV expresses a handful of immunomodulatory (BCRF1, BARF1, BNLF2A, BGLF5 & BILF1) and anti-apoptotic (BHRF1 & BALF1) proteins. In this review, we discuss the evidence supporting an abortive lytic cycle in which these lytic genes are expressed, and how the immunomodulatory mechanisms of EBV and related herpesviruses Kaposi Sarcoma herpesvirus (KSHV) and human cytomegalovirus (HCMV) result in paracrine signals that feed tumor cells. An abortive lytic cycle would reconcile the need of lytic expression for viral tumorigenesis without relaying in a complete cycle that would induce cell lysis to release the newly formed infective viral particles.