Over the last decades, attempts to develop new therapies for acute heart failure (AHF) have largely failed. Limitations in understanding the pathophysiology of AHF, its natural history, the effects of current therapies, the properties of new agents, and, importantly, study designs and execution have contributed to these disappointing results. We propose a novel approach for the development of new agents for AHF. Initial stages of development (proof of activity and phase IIa studies) should provide a granular understanding of the mechanism of the actions and pharmacodynamic properties of the new intervention, first in animal models of HF and then in small tightly designed human studies. This will facilitate the identification of patients who are more likely to show a favorable response. In phase II studies (proof of concept), the assessment of efficacy and final dose selection should not be based on surrogate endpoints but rather on clinically meaningful endpoints such as the relief of AHF symptoms, i.e. breathlessness, as well as consistent trends in the relief of other symptoms and clinical signs of recurrence (worsening HF either in hospital or after discharge) and short-term mortality. Lastly, other safety parameters, i.e. effects such as renal and myocardial damage, should be assessed. In some instances, the decision to move to phase III may be based on clear consistent trends rather than on 1 statistically significant endpoint. In studies that confirm clinical utility (phase III), efficacy will be judged relative to the magnitude of the benefit, i.e. improvements in symptoms (breathlessness), the prevention of symptom recurrence, and short-term mortality, balanced by the consideration of safety signals including an estimation of absolute mortality.