The approval of new heart failure (HF) therapies has slowed over the past two decades in part due to the high costs of conducting large randomized clinical trials that are needed to adequately power major clinical endpoint studies. Several biomarkers have been identified reflecting different elements of HF pathophysiology, with possible applications in diagnosis, risk stratification, treatment monitoring, and even in the design of clinical trials. Biomarkers could potentially be used to refine study inclusion criteria to enable enrolment of patients who are more likely to respond to a therapeutic intervention, despite being at sufficient risk to meet pre‐determined study endpoint rates. When there is a close relationship between biomarker levels and clinical endpoints, changes in biomarker levels after a given treatment can act as a surrogate endpoint, potentially reducing the duration and cost of a clinical trial. Natriuretic peptides have been widely used in clinical trials with a variable amount of added value, which such variation being probably due to the absence of a close pathophysiological connection to the study drug. Notable exceptions to this include sacubitril/valsartan and vericiguat. Future studies should seek to adopt unbiased approaches for discovery of true companion diagnostics; with ‐omics‐based tools, biomarkers might be more precisely selected for use in clinical trials to identify responses that closely reflect the biological effects of the drug under investigation. Finally, biomarkers associated with cardiac damage and remodelling, such as cardiac troponin, could be employed as safety endpoints provided that standardization between different assays is achieved.