The overall purpose of this study was to evaluate the pharmacodynamic response to warfarin in cats. The specific aim was to determine if a log‐linear indirect response model (Nagashima et al., 1969) used to describe the in vivo effect of warfarin in humans could be applied to cats. The pharmacokinetics of racemic warfarin were described using a non‐compartmental approach. The relationship between prothrombin complex activity (PCA) and normalized prothrombin time (PTR) was defined for feline plasma under our experimental conditions, and determined to be: %PCA=12.38+648 e−PTR/0.492. These data were then integrated and used to predict the warfarin dose associated with therapeutic anti‐coagulation defined as an International Normalized Ratio (INR) of 2.0–3.0.The maximum prothrombinopenic response to warfarin in cats after a single intravenous dose of 0.5 mg/kg occurred at 24–48 h. Pharmacodynamic modeling suggested that each cat had a narrow therapeutic range of the steady‐state concentration of total warfarin required to appropriately block prothrombin complex synthesis (median: 265.2–358.7 ng/mL). The median daily dose range predicted to yield therapeutic concentrations of warfarin was 0.061–0.088 mg/kg per day. Wide inter‐individual variations in both pharmacokinetics and pharmacodynamic response suggest that a more optimal dosing of warfarin may be possible with the development of individual pharmacokinetic/pharmacodynamic algorithms, analogous to those currently employed in human patients.