A subclinical vitamin K deficiency was induced in 32 healthy subjects (four groups of eight males and females) aged 20-40 and 60-80 yr residing in the Metabolic Research Unit of the Human Nutrition Research Center on Aging at Tufts University. Volunteers were initially fed (4 d) a baseline-period diet containing the recommended daily allowance for vitamin K which is equivalent to 80 ;tg/d of phylloquinone (vitamin K1).During the baseline period various parameters of vitamin K nutritional status were monitored. The baseline period was followed by a 13-d depletion period during which the subjects were fed a very low vitamin K1 diet (-10 ;ug/d). After depletion, the subjects entered a 16-d repletion period (four stages lasting 4 d each) during which time they were repleted with 5, 15, 25, and 45 ,ug of vitamin K1 per day. Vitamin K1 depletion dramatically and significantly decreased plasma vitamin K1 levels (P < 0.0001) in both elderly and young groups to values 13-18% of day 1 (elderly 0.22 nM, young 0.14 nM). Repleting the subjects with up to 45 ;ig of vitamin K1 per day failed, in the case of the young subjects, to bring plasma vitamin K1 levels back into the normal range. Dietary vitamin K1 restriction induced different responses in the urinary excretion of-y-carboxyglutamic acid between the young and the elderly subjects with values decreasing significantly (P < 0.03) in the young while remaining unchanged in the elderly. The vitamin K1 depletion period had no significant effect on either prothrombin and activated partial thromboplastin times, or Factor VII and protein C (as determined by antigenic and functional assays). By using a monoclonal antibody, descarboxy prothrombin was found to increase slightly but significantly in both groups (P < 0.05) as a consequence of the low vitamin K1 diet. This study clearly shows that a diet low in vitamin K1 can result in a functional subclinical deficiency of vitamin K (decreased urinary y-carboxyglutamic acid excretion) without affecting blood coagulation.