The disposition of a single intravenous bolus dose of 10 mg vitamin K1 and vitamin K1-2,3-epoxide were studied in two healthy subjects without and with 12 h pretreatment dose of phenprocoumon (0.4 mg/kg). For each compound administered alone the plasma concentration-time profile was adequately fitted by a biexponential equation, with an average terminal half-life of 2.0 and 1.15 h for the administered vitamin K and its 2,3-epoxide respectively. While vitamin K1 was measurable in plasma following administration of vitamin K1-2,3-epoxide, the epoxide was not detectable following administration of vitamin K1. Following pretreatment with phenprocoumon and after intravenous administration of vitamin K1, both the average half-life and area under the plasma concentration-time profile of vitamin K1 were marginally reduced to 1.5 h and 1.76 mg l-1 h respectively, while the plasma concentration of vitamin K1-2,3-epoxide was readily measurable and its half-life markedly prolonged to 14.7 h. Following pretreatment with phenprocoumon and after oral administration of vitamin K1-2,3-epoxide, no vitamin K1 was detectable in plasma and the half-life of the epoxide was 13.8 h. Based on area considerations the data suggest that either phenprocoumon does more than just inhibit the reduction of vitamin K1-2,3-epoxide to vitamin K1, or that the simple model describing the interconversion between vitamin K1 and its epoxide is inadequate. The same conclusion is drawn from the analysis of comparable data in dogs, obtained by Carlisle & Blaschke (1981).