Public funding for research on the action of drugs in countries like the U.S. requires that racial classification of research subjects should be considered when defining the composition of the samples as well as in data analysis. In a controversial arena, once race is included in research designs, it is created the possibility of interpreting that whites and blacks are so distinct that pharmacogenes present in the genetic background of black people would be absent from the genetic background of white and vice versa. In Brazil, pharmacogenomic results have led to very different interpretations when compared to those obtained in U.S. This is explained as deriving from the genomic heterogeneity of the Brazilian population. The premise is that admixture prevents the direct association between color and genetic ancestry. This paper addresses the relationship between scientific practice and the naturalization of social values in biomedical research. Our data derive from anthropological investigation conducted in INCA (Brazilian National Cancer Institute) with a focus on the drug warfarin, which are compared with similar data derived from research carried out in the US and Europe. The criticism of Brazilian scientists regarding the uses of racial categorization includes a revision of mathematical algorithms for drug dosage widely used in clinical procedures around the world. Our analysis reveals how the incorporation of ideas of racial purity and admixture, as it relates to the efficacy of drugs, touches on issues related to the possibility of application of pharmaceutical technologies on a global scale.