The aim of our study was to investigate the effect of a diet with a known vitamin K content on the monitor test for oral anticoagulant therapy. We studied 10 poorly controlled patients (7 women and 3 men, mean age 48 +/- 15 years) in therapy with acenocoumarol for at least 1 year. Another group of 10 poorly controlled patients was considered as a control group. After a baseline period, during which a free diet was allowed, patients were administered a diet with a controlled vitamin K content; in the same period, control subjects were not subjected to any dietary restriction. Thrombotest (Nyegaard, Oslo) was employed for monitoring oral anticoagulant therapy, with a therapeutic range between 2.3 and 4.8 INR. We found a significant difference by thrombotest between the percentages within the therapeutic range obtained before and during dietary treatment (p = 0.0001). The difference in percentages was -0.31 and the 95% CI, of the difference ranged from -0.45 to -0.17. No significant difference was shown in the control group in the same periods. Our data suggest that a diet with a controlled vitamin K content is effective in increasing the percentage of tests within the therapeutic range in patients with poorly controlled anticoagulation.