AbstractPredictive biomarkers play an important role in our efforts to individualize pharmacotherapy, and within recent years, a number of different types of assays have been introduced. These biomarkers may potentially support the selection and dosage of specific drugs in order to maximize efficacy and minimize adverse reactions in the individual patient. However, in many instances, the scientific and clinical evidence is insufficient to support the prescribing decision. When predictive biomarkers are used to guide pharmacotherapy, it is important to secure that decisions are based on solid clinical evidence. Here, the regulatory authorities, especially the FDA, have been at the forefront in relation to regulate this type of biomarker assay in order to secure patient safety. The approval process for companion diagnostics is an example of this effort, where the scientific validity of the biomarker and assay is in focus. With the approaching implementation of the new IVD Regulation, greater attention will also be paid to analytical and clinical validity of biomarker assays in the EU. For any type of predictive biomarker assay, including pharmacogenetic and tumour profiling tests, the clinical evidence needs to be in place before they are used routinely in the clinic.