In summary, dose-response relationships have been demonstrated for Ni3S2 carcinogenesis in rats and hamsters and for transformation of Syrian hamster fetal cells by Ni3S2 in vitro. Absolute species specificity has not been observed in Ni3S2 carcinogenesis, although rats are apparently more susceptible than mice, hamsters, or rabbits. Also, significant variations have been reported in susceptibilities of rat strains to Ni3S2 carcinogenesis. Most organs of rats have been found to be susceptible to Ni3S2 carcinogenesis following direct exposure by injection or inhalation; intraocular and intramuscular routes of administration have yielded the highest tumor incidences. Finally, an experiment in hamsters has indicated that Ni3S2 may be noncarcinogenic by the oral route; further studies are needed to confirm or refute this speculation. For discussions of molecular mechanisms that may be involved in carcinogenesis by Ni3S2 and other nickel compounds, readers are referred to recent review articles (8-10).