The refinement of disease taxonomy utilizing molecular phenotypes has led to significant improvements in the precision of disease diagnosis and customization of treatment options. This has also spurred efforts to identify novel biomarkers to understand the impact of therapeutically altering the underlying molecular network on disease course, and to support decision-making in drug discovery and development. However, gaps in knowledge regarding disease heterogeneity, combined with the inadequacies of surrogate disease model systems, make it challenging to demonstrate the unequivocal association of molecular and physiological biomarkers to disease pathology. This article will discuss the current landscape in biomarker research and highlight strategies being adopted to increase the likelihood of transitioning biomarkers from discovery to medical practice to enable more objective decision making, and to improve health outcome.