Genetic markers displaying highly significant statistical associations with complex phenotypes may not necessarily possess sufficient clinical validity to be useful. Understanding the contribution of these markers beyond readily available clinical biomarkers is particularly important in pharmacogenetics. We demonstrate the utility of genetic testing using the example of warfarin in a multi-ethnic setting comprising of three Asian populations that are broadly representative of the genetic diversity for half of the population in the world, especially as distinct interethnic differences in warfarin dose requirements have been previously established. We confirmed the roles of three well-established loci (CYP2C9, VKORC1 and CYP4F2) in explaining warfarin dosage variation in the three Asian populations. In addition, we assessed the relationship between ethnicity and the genotypes of these loci, observing strong correlations at VKORC1 and CYP4F2. Subsequently, we established the additional utility of these genetic factors in predicting warfarin dose beyond ethnicity and clinical biomarkers through performing a series of systematic cross-validation analyses of the relative predictive accuracies of various fixed-dose regimen, clinical and genetic models. Through a pharmacogenetics model for warfarin, we show the importance of genetic testing beyond readily available clinical biomarkers in predicting dose requirements, confirming the role of genetic profiling in personalized medicine.