OBJECTIVES: Pharmacogenomic biomarkers aid in predicting drug response and adverse drug reactions. Drug label provides information about these biomarkers; however the quality of evidence regarding biomarker use is unclear. Thus, we investigated the FDA-approved drug labels for the availability and quality of evidence supporting the biomarkers use in conjunction with targeted therapies in different cancers. METHODS: We searched the US FDA website to identify the labels providing information on the pharmacogenomic biomarkers of the targeted cancer therapies. The Evaluation of Genomic Application in Practice and Prevention Working Group’s guideline was used to assess the clinical validity and utility of the referenced studies. The available evidence was graded as convincing, adequate, or incomplete. We also assessed the completeness of studies and recommendation in the label. RESULTS: Fifty-three drug-biomarker combinations were identified, encompassing 42 unique drugs and 23 unique biomarkers. Combinations were most frequently identified in breast cancer (26%), chronic myeloid leukemia (15%), and colorectal cancer (11%). Half of the supporting evidence in drug labels (51%) were not graded with convincing validity (i.e., the ability to predict the phenotype of interest) and more than half (60%) were incomplete pertaining to utility (i.e., the ability to improve measurable clinical outcomes). Complete information of the referenced clinical studies was included in only 11% labels, whereas 62% labels provided partial information. The treatment recommendations about clinical decision were based on the drug’s mechanism of action in 75% of the labels and on drug-biomarker association in 21% of drug labels. Also, the biomarker’s prediction of improved drug response and contraindications was provided in 77% and 23% labels, respectively. CONCLUSIONS: Majority of the labels lacked convincing validity and utility regarding biomarker use. As biomarkers may enhance clinical care, it has become extremely important for inclusion and rational use of pharmacogenomic information in drug labels, for optimized decision-making.