A novel generation of drugs is introduced in the treatment of heart failure (HF). These drugs, including phosphodiesterase-5 inhibitors, guanylate cyclase stimulators and activators, share the feature that their action is either endothelial-mediated or substitutes for endothelial pathways, in particular the nitric oxide-cyclic guanosine monophosphate pathway, thereby influencing homeostatic balances in virtually each organ system in a pleiotropic fashion. Unfortunately, recent clinical trials with some of these drugs have shown disappointing results, at least in the setting of HF with a preserved ejection fraction. This suggests that their clinical use may require approaches that diverge from traditional pharmacological approaches, the latter often titrated on the effects of drugs on haemodynamic parameters or single biomarkers. In this paper we preconize that HF drugs with an endothelial profile should be applied conform to principles of endothelial physiology and systems pharmacology. This type of drug therapy should be viewed as a systems physio-pharmacological intervention and its clinical use accustomed to systems pharmacological principles, comparable to the systemic endothelial-mediated benefits induced by exercise training in HF. We will review the actions of these drugs and define criteria to which trials with these drugs should comply in order to increase chances of success.