Aim of the study was to compare numbers of episodes of excess hypocoagulation and bleeding with warfarin dosing based on pharmacogenetic testing and traditional method in patients with high risk of thromboembolic complications. In 76 patients (43 men and 33 women aged 60.3 +/- 12.3 years) warfarin was administered starting with the dose calculated according to the gage algorithm with consideration of results of pharmacogenomic testing (genotyping of CYP2C9 and VKORC1). Control group comprised 78 patients aged 63.4 +/- 9.4 years who had participated in an earlier retrospective study in which they received warfarin according to traditional scheme with starting dose of 5 mg/day. In both groups we analyzed data obtained during 6 months after start of drug administration. Genotyping was carried out by polymerase chain reaction. Episodes of excess hypocoagulation (international normalized ratio above therapeutic range) and bleeding accurred more rarely with the use of pharmacogenetic approach to dosing of warfarin compared with standard method (17.1 vs 56.4%, p = 4.1 x10(-7), and 4 vs 18%, p = 0.009 respectively).