Male B6C3F1 mice and male F344/N rats were exposed to chloral hydrate (chloral) in the drinking water for 2 years. Rats: Measured chloral hydrate drinking water concentrations for the study were 0.12 g/L, 0.58 g/L, and 2.51 g/L chloral hydrate that yielded time-weighted mean daily doses (MDDs) of 7.4, 37.4, and 162.6 mg/kg per day. Water consumptions, survival, body weights, and organ weights were not altered in any of the chloral hydrate treatments. Life-time exposures to chloral hydrate failed to increase the prevalence (percentage of animals with a tumor) or the multiplicity (tumors/animal) of hepatocellular neoplasia. Chloral hydrate did not increase the prevalence of neoplasia at any other organ site. Mice: Measured chloral hydrate drinking water concentrations for the study were 0.12 g/L, 0.58 g/L, and 1.28 g/L that gave MDDs of 13.5, 65.0, and 146.6 mg/kg per day. Water consumptions, survival, body and organ weights, were not altered from the control values by any of the chloral hydrate treatments. Enhanced neoplasia was observed only in the liver. Prevalence and multiplicity of hepatocellular carcinoma (HC) were increased only for the high-dose group (84.4%; 0.72 HC/animal; p < or = 0.05). Values of 54.3%; 0.72 HC/animal and 59%; 1.03 HC/animal were observed for the 13.5- and 65.0-mg/kg per day treatment groups. Prevalence and multiplicity for the control group were 54.8%; 0.74 HC/animal. Hepatoadenoma (HA) prevalence and multiplicity were significantly increased (p < or = 0.05) at all chloral hydrate concentrations: 43.5%; 0.65 HA/animal, 51.3%; 0.95 HA/animal and 50%; 0.72 HA/animal at 13.5, 65.0, and 146.6 mg/kg per day chloral hydrate compared to 21.4%; 0.21 HA/animal in the untreated group. Altered foci of cells were evident in all doses tested in the mouse, but no significant differences were observed over the control values. Hepatocellular necrosis was minimal and did not exceed that seen in untreated rats and mice. Chloral hydrate exposure did not alter serum chemistry and hepatocyte proliferation in rats and mice or increase hepatic palmitoyl CoA oxidase in mice at any of the time periods monitored. It was concluded that chloral hydrate was carcinogenic (hepatocellular neoplasia) in the male mouse, but not in the rat, following a lifetime exposure in the drinking water. Based upon the increased HA and combined tumors at all chloral hydrate doses tested, a no observed adverse effect level was not determined.