The human vitamin K requirement is not known precisely, but the minimal requirement is often assumed to be between 0.5 and 1 x 10(-6) g/kg body weight. In the present study we addressed the question to what extent circulating vitamin K concentrations are influenced by the form in which the vitamer is consumed. The experimental group consisted of five healthy volunteers who received phylloquinone after an overnight fast. On the first day of three successive weeks the participants consumed 1 mg (2.2 mumol) phylloquinone, either in the form of a pharmaceutical preparation (Konakion), or in the form of spinach + butter, or as spinach without added fat. Circulating phylloquinone levels after spinach with and without butter were substantially lower (7.5- and 24.3-fold respectively) than those after taking the pharmaceutical concentrate. Moreover, the absorption of phylloquinone from the vegetables was 1.5 times slower than from Konakion. In a second experiment in the same five volunteers it was shown that relatively high amounts of menaquinone-4 enter the circulation after the consumption of butter enriched with this vitamer. It is concluded that the bioavailability of membrane-bound phylloquinone is extremely poor and may depend on other food components, notably fat. The bioavailability of dietary vitamin K (phylloquinone + menaquinones) is lower than generally assumed, and depends on the form in which the vitamin is ingested. These new insights may lead to a revision of the recommended daily intake for vitamin K.