Current studies have shown that all mammalian sarcoma-producing viruses, whether isolated from laboratory experiments of naturally occurring tumors, are deletion mutants of replicating mammalian type C viruses. Nonproducer cells transformed by any of these sarcoma viruses contain RNA homologous to mammalian leukemia viruses, even though the cells contain no known proteins currently coded for by the mammalian leukemia virus. This mammalian leukemia virus information (FT-) is a genetically stable part of the mammalian sarcoma viruses (FT+). Second, another component in the Kirsten and Harvey sarcoma viruses can be identified in addition to this leukemia virus information for the homologous leukemia virus; at least part of the additional information came from rat type C viruses from the animals in which the sarcoma viruses were isolated. This indicates that these two mammalian sarcoma viruses are recombinants between mouse leukemia virus and genetic information in rat cells and suggests that the process of formation of the sarcoma virus is analogous to transduction of information in bacteriophage. Third, the Kirsten sarcoma virus seems to have a third component in it separate from either the mouse leukemia virus or rat leukemia virus information. Fourth, and FT+ leukemia virus isolated from mice, the Abelson leukemia virus which causes as B-cell leukemia, is also defective and can be shown to have information homologous to Moloney leukemia virus. Fifth, in the feline sarcoma virus, feline leukemia information can be detected, but information for the other cat virus, RD-114, cannot be detected. Finally, mutants of Kirsten sarcoma virus temperature sensitive for the maintenance of transformation have been isolated, and out of ten such mutants, thus far no complementation has been observed.