In the absence of adequate data on humans, it is biologically plausible and prudent to regard agents and mixtures for which there is sufficient evidence of carcinogenicity in experimental animals, usually rats and mice, as if they presented a carcinogenic risk to humans. Prediction of cancer sites in humans from bioassay data in rodents is much less certain, however, regardless of organ or tissue. For tumors of the nervous system, there is practically no basis for judging the validity of such predictions, as only ionizing radiation is known to cause tumors of the central nervous system (CNS) in humans. Brain tumors are relatively uncommon findings in bioassays and are rare in untreated rodents, even in rats, which appear to be the most susceptible species. However, CNS tumors have been readily induced in rodents by systemic exposures to some chemicals, notably N-nitrosoalkylureas and other alkylating agents and certain alkyl hydrazine derivatives. CNS tumors in rodents have played a significant role in carcinogenic hazard evaluations of several other chemicals, including acrylonitrile, ethylene oxide, and acrylamide, and have been implicated as part of the tumor spectrum induced by vinyl chloride and certain inorganic lead compounds. In some of these evaluations, it is not certain that all tumors diagnosed as primary brain tumors were correctly identified. Diagnostic difficulties have been presented by undifferentiated small-cell tumors that may invade the brain, including carcinomas of the nasal cavity and undifferentiated schwannomas arising in cranial nerve ganglia, and by the difficulty of reliably distinguishing between focal reactive gliosis and early glial neoplasms. The most striking experimental finding regarding the induction by chemicals of tumors of the nervous system is the dramatically greater susceptibility of the fetal and neonatal nervous system to some carcinogens, as compared with the susceptibility of the nervous system in adults of the same species.