It is unclear what advice should be given to patients using vitamin K antagonists with respect to dietary vitamin K intake. We performed a nested case-control study in patients attending a Dutch anticoagulation clinic, to determine the association between vitamin K intake and subtherapeutic International Normalized Ratio (INR) values and the interaction between usual and recent vitamin K intake. Compared to patients with a normal usual vitamin K intake, those with a high vitamin K intake had a decreased risk of a subtherapeutic INR [Hazard Ratio (HR) 0.80, 95% confidence interval (CI): 0.56-1.16] and patients with a low vitamin K intake an increased risk (HR 1.33, 95% CI: 0.79-2.25). In patients with a low usual vitamin K intake, recent vitamin K intake was twice as high in cases as in controls (164 vs. 85 microg/d). Such a difference was not observed in patients with a normal or high usual vitamin K intake. Our results suggest that a high vitamin K intake reduces the risk of a low INR by lessening the influence of incidental consumption of vitamin K rich food on the INR. These findings support the recommendation for patients on vitamin K antagonists to eat a sufficient amount of vitamin-K containing foods.