Patients with cardioembolic stroke and stroke due to dissection of extracranial artery or cerebral sinus thrombosis require urgent anticoagulant therapy initiation.This high risk group of patients are at risk of 5-7% of embolic stroke recurrence within the first week or at the risk of bleeding into the infarct zone with worsening of neurologic deficits. Many reports confirmed that CYP2C9 and VKORC1 genetic polymorphisms have strong influence on interindividual warfarin sensitivity, variability of anticoagulant effects and dose requirement.Therefore, in patients with ischemic stroke it is important to select initial doses of warfarin based on individual genotype.The aim of individualisation is to achieve earlier anticoagulant effect and stable maintenance dose in order to reduce the risk of developing warfarin therapy side effects in early initiation of therapy due to hypo or overanticoagulation, particularly among carriers of multiple allelic variants. At the Department of neurology University hospital Centre Zagreb we conducted a prospective case-control study during 6 months among patients hospitalized for acute ischemic stroke with indications for anticoagulant therapy. The hypothesis was that by using the algorithm for pharmacogenetic polymorphism of CYP2C9 gene (wt, * 2, * 3 ) and VKORC1 1173C> T in these patients before initiating warfarin therapy, one can achieve an erlier anticoagulant effect and a stable maintenance dose in comparasion to the introduction of a standard fixed dose. The consented stroke patients (n=106) admitted at Referral centre for intensive neurologic care were pharmacogenetic tested for CYP2C9 and VKORC1 polymorphism before the initial dose of warfarin was predicted. The control group (n=104) were consented acute stroke patients hospitalized at other subunits of Department of neurology with the drug induction using the standard fixed dose, without pharmacogenetic prediction. Among both groups of patients we recorded initial International normalised ratio (INR), INR after 48 hours, 72 hours, on day 5., 7.,14. and 21. of warfarin therapy induction and studied time to reach first target INR value, time to reach stable maintance dose, percent of time spent in therapeutic INR range and appearance of eventual complications. The main goal of this research is to achieve the first target INR erlier, to have larger proportion of time spent in therapeutic INR range and to reach earlier achievement of a stable maintenance dose. Mentioned goals were achieved and confirmed by this study among patients with initial dose introduction after pharmacogenetic algorithm (FG) in comparasion with the group to which standard fixed dose was given. Time spent in over therpeutic INR >3,1 was shorter in genotype guided group, compared to standard fixed dose initiating group. Application of pharmacogenetic algorithm enabled the correct assessment of warfarin dose in 81,5% of 62% of patients who required a higher or lower dose of warfarin. We found no reduction in the development of complications in the FG group compared to group with standard fixed dose application, but we have proved the difference between clinical outcome in this group patients with complications had better clinical outcome due to the occurrence of mild bleeding in the FG group, in comparison with severe bleeding in patients using a standard fixed dose.We've shown a better clinical outcome with mild neurological deficits in relation to neurological deficit on admission among patients with pharmacogenetic prediction of doses in comparasion to the group of patients with standard fixed doses. Our findings highlight the importance of selection of initial warfarin dose based on individual genotype to therapy instead of a fixed dose for safer therapeutic intervention in high risk patients with acute stroke.