Regulatory changes in the use of some second-generation anticoagulant rodenticides in parts of North America may result in expanded use of first-generation anticoagulant rodenticides (FGARs). Recent toxicological studies with captive raptors have demonstrated that these species are considerably more sensitive to the FGAR diphacinone than traditional avian wildlife test species (mallard, bobwhite). We have now examined the toxicity of the FGAR chlorophacinone (CPN) to American kestrels fed rat tissue mechanically-amended with CPN, or rat tissue containing biologically-incorporated CPN, for 7 days. Nominal CPN concentrations in these diets were 0.15, 0.75 and 1.5 µg/g food wet weight, and actual CPN concentration in diets were analytically verified as being close to target values. Food intake was consistent among groups, body weight fluctuated by less than 6%, exposure and adverse effects were generally dose-dependent, and there were no dramatic differences in toxicity between mechanically-amended and biologically-incorporated CPN diets. Using benchmark dose statistical methods, toxic reference values at which clotting times were prolonged in 50% of the kestrels was estimated to be about 80 µg CPN consumed/kg body weight-day for prothrombin time and 40 µg CPN/kg body weight-day for Russell's viper venom time. Based upon carcass CPN residues reported in rodents from field baiting studies, empirical measures of food consumption in kestrels, and dietary-based toxic reference values derived from the 7-day exposure scenario, some free-ranging raptors consuming CPN exposed prey might exhibit coagulopathy and hemorrhage. These sublethal responses associated with exposure to environmentally realistic concentrations of CPN could compromise survival of exposed birds.