Basic and therapeutic trial results obtained in the spontaneous AK leukemia (lymphoma) model have been brought together for comparison with available information on the much used transplanted murine leukemia models and human leukemias and lymphomas. The etiologic agent for "spontaneous" AK lymphoma is an RNA virus present at birth in AKR mice. Lymphoma cells first appear in the thymuses of animals at 5-->12 months of age. The time lapse between the first appearance of viable lymphoma cells in the thymus and clinical diagnosis (eg, with about 10(9) widely disseminated viable plus nonviable lymphoma cells in the host) is about 1 month. Thus, the overall doubling time of lymphoma cells before diagnosis is about 1 day. This estimate is compatible with the doubling time of relatively small numbers of first-passage lymphoma cells, assay data on the rate of repopulation of viable lymphoma cells after therapeutic reduction, and the median intermitotic time of dividing lymphoma cells (ie, 0.6 day). In general, the cytokinetic parameters of advanced spontaneous AK lymphoma cell populations are more like those observed in advanced human leukemias than are those of early L1210 leukemia. This paper presents assay data on the reduction in viable spontaneous AK lymphoma cells after treatment with a variety of agents, and the rate of cell repopulation after cessation of treatment. Extensive therapeutic trial data indicate that cyclophosphamide is presently the most effective single agent against spontaneous AK lymphoma, with arabinosylcytosine or palmO-ara-C a close second. Daunomycin, 5-fluorouracil, the nitrosoureas, vincristine, methotrexate, and dexamethasone provided moderate increases in host survival time. The combination of vincristine plus prednisone was a good remission inducer but the median survival time after cessation of treatment was shorter than that observed for cyclophosphamide or palmO-araC. The best responses observed to date with two-drug combinations appear better on several scores than the best that have been observed with single drugs. The best overall responses observed to date with two-drug combinations were with palmO-ara-C plus methyl-CCNU, cyclophosphamide plus methyl-CCNU, and palmO-ara-C plus cyclophosphamide. Some three- and four-drug combinations have provided better therapeutic responses than have been observed with single agents but not significantly better than those obtained with two-drug combinations. Splenomegaly assays carried out immediately after cessation of treatment and 60 days and longer after cessation of treatment, suggest that eradication of all viable lymphoma cells is being achieved in some animals by combination chemotherapy; however, such animals eventually die of lymphoma, presumably as a result of the reinduction of a second lymphoma cell population. The requirements for permanent "cure" of spontaneous lymphoma in AKR mice include eradication of all viable lymphoma cells and prevention of reinduction. Two major differences between early L1210 leukemia and clinically diagnosed spontaneous AK lymphoma are the degree of disease advancement at the time therapy is usually started (and associated cytokinetic differences) and the reinduction problem in AKR mice. Spontaneous AK lymphoma is relatively more advanced at diagnosis than is acute leukemia in man (ie, with respect to nearness of the host to death), and it is presumed that the reinduction problem in AKR mice is more acute and more prevalent than in human neoplastic disease.