The YT cells, already known as natural killer (NK) line, were tested for Epstein-Barr virus (EBV). The simply maintained YT cells (YT-O) and the two different subclones showed and identical length of junctional DNA of terminal repeats in Southern blot with LMP-1 probe, indicating that the 3 had already been positive for EBV before subcloning. YT-O expressed limited amount of EBNA2 or LMP-1 mRNA, whereas the 2 subclones expressed abundant EBNA2 or LMP-1 mRNA in the Northern blot analysis with EBNA2 or LMP-1 probe. Thus, the ex vivo cells were positive for EBV, since the BL (Burkitt lymphoma) type EBV gene expression in the YT-O does not generally occur in in vitro infection. The remaining clinical record indicated that the lymphoma was extranodal (angiocentric lymphoma), involving mediastinum and liver, but not nodal or lymphoblastic lymphoma (LBL). Acute lymphoblastic lymphoma (ALL)/LBL of the NK-lineage has not been defined, although such neoplasms should exist. Since T- and NK-lineages are so close in immature stages of differentiation that ALL/LBL of NK may have been sorted into T-lineage. The phenotypic records of the T-ALL/LBL in our laboratory indicated that CD7+CD5+CD2- is much higher in incidence than CD7+CD5-CD2+. This may reflect the size difference of physiological populations of T- and NK-lineage cells. Furthermore, the latter is of CD45RO type in contrast to the rest of the early-thymic (pro-thymic) T-ALL/LBL groups of CD45RA type. A CD56+ case of 4 CD7+CD5-CD2+ cases have been published as a case of LBL of NK-lineage. It is necessary to scrutinize CD7+CD5-CD2+ cells in order to clarify the phenotype of neoplastic and physiological NK cells in immature stages.