This review focuses on the cellular origin of Hodgkin and Reed-Sternberg (HRS) cells, their association with the Epstein-Barr virus (EBV), and their relation to Ki-1+ anaplastic large-cell (ALC) lymphoma. The tingibility of HRS cells in paraffin sections for polyclonal immunoglobulin represents a staining artifact and thus can no longer serve as an argument for the histiocytic nature of HRS cells. Immunolabeling studies do not support the putative relationship of HRS cells to cell types such as macrophages or interdigitating reticulum cells, but instead suggest: a) that lymphocyte-predominant (LP) Hodgkin's disease (HD) represents a B-cell neoplasm which is distinct from non-LP HD, and b) that non-LP HD constitutes a syndrome rather than a disease entity, with the existence of T-cell types and B-cell types. HRS cells (and the tumor cells in ALC lymphomas) frequently display an immature genotype in association with late activation markers, leading to the assumption that the tumor cells in many cases of HD (and some cases of ALC lymphoma) may be derived from immature lymphoid cells that are infected by a virus that superimposes characteristics of mature activated lymphocytes on these cells. Southern blotting, in situ hybridization, and polymerase chain reaction (PCR) experiments revealed an association of EBV with HRS cells in a significant proportion of HD cases, suggesting that EBV may be responsible for the dissociation between genotype and phenotype in HRS cells, because EBV is a strong inducer of the activation antigens CD30 and CDw70.(ABSTRACT TRUNCATED AT 250 WORDS)