Despite some exciting new leads, human immunodeficiency virus-I (HIV-I)-related non-Hodgkin's lymphoma (HIV-NHL) remains a fatal malignancy for the vast majority of patients. The use of highly active antiretroviral therapy (HAART) has not produced a fall in the incidence of HIV-NHL and conventional cytotoxic chemotherapy is associated with a negligible cure rate. New treatment options are needed. Future therapeutic directions in HIV-NHL should be based on a better understanding of three fundamental aspects of lymphomagenesis in the setting of acquired immunodeficiency. (I) New information on the molecular and pathological heterogeneity of HIV-NHL should be applied to the development of risk-adapted therapy. The identification of patient subsets with different susceptibility to cytotoxic chemotherapy, immunomodulation, or antiviral strategies is essential for the design of clinical trials of investigational new agents in HIV-NHL. (2) Known viral pathogens need to be better understood. The presence and role of new viruses should be investigated. Key biological interactions between virus and host, mediated by oncogenic, immunomodulatory, and antiapoptotic viral proteins, should become the main target for new drug development. (3) Immune reconstitution with HAART and immunostimulatory cytokines such as interleukin-2 (IL-2) and IL-12, combined with drugs that downregulate the replication or gene expression of tumor-associated viruses such as Epstein-Barr virus (EBV) and human herpes virus-8 (HHV-8), possibly in combination, should remain a primary goal in the treatment of HIV-NHL. Preventive immunization, using autologous tumor cell vaccines or antigen-specific cellular adoptive immunotherapy should be explored.