Plasma volume expansion in patients with heart failure (HF) is associated with neurohumeral activation; symptom and signs of pulmonary congestion and peripheral oedema; an increase in hospitalizations for HF; and cardiovascular mortality. A reduction in plasma volume, as evidenced by haemoconcentration, has recently been suggested to be associated with an improvement in cardiovascular mortality. However, while loop diuretics reduce plasma volume and relieve the symptoms and signs of pulmonary congestion, they activate the renin–angiotensin–aldosterone system (RAAS) and, in experimental studies and some, although not all, retrospective clinical studies have been associated with an increase in cardiovascular mortality. Due to the controversy associated with the use of high-dose loop diuretics and their effect on cardiovascular mortality in patients with HF and a reduced left ventricular ejection fraction (REF) there has been increasing interest in alternative and or supplemental diuretic strategies to reduce plasma volume and to improve cardiovascular outcomes in patients with HFREF. Tolvaptan, a selective V2 vasopressin antagonist, has been shown in pre-clinical and short-term studies in patients with HF to be a potent aquaretic with a resultant increase in urine output; a decrease in body weight; a decrease in pulmonary capillary wedge pressure; an increase in serum sodium concentration; without activation of the RAAS. However when studied over the longer term in patients with HFREF, it has not been shown to reduce natriuretic peptide levels or left ventricular remodelling, and did not improve cardiovascular outcomes despite an early improvement in symptoms of dyspnoea, a prolonged reduction in body weight, and normalization of serum sodium in hyponatraemic patients. The finding that tolvaptan failed to reduce natriuretic peptide levels and cardiovascular outcomes despite a reduction in body weight is especially notable since an increase in body weight post-discharge is the single most important predictor for readmission in patients with heart failure and a major predictor of mortality. The explanation for the failure of tolvaptan to improve ventricular remodelling, natriuretic peptide levels, and cardiovascular outcomes despite a reduction in body weight remains uncertain. Several explanations have, however, been proposed, including the suggestion that the dose of tolvaptan (30 mg/day) may have been inadequate; that patients selected for study did not have hyponatraemia; and the possibility that a reactive increase in vasopressin levels might over the long term stimulate the V1a receptor with a resultant increase in vasoconstriction, which might negate the beneficial effects associated with aquareses. Without further prospective mechanistic studies it is difficult to confirm or refute these explanations. Recently, however, it has been shown that a deficiency of the V1a receptor causes hyporeninaemic hypoaldsteronism, and stimulation of the V1a receptor causes the adrenal production of aldosterone. The V1a receptor also affects the renal tubular effects of aldosterone. We postulate that these findings may explain the paradox of a persistent reduction in body weight despite the lack of an improvement in natriuretic peptide levels and cardiovascular outcomes associated with the use of tolvaptan in EVEREST. This hypotheses and its implications for the therapy of patients with HF will be briefly discussed below. As noted above, V2 receptor antagonism results in a sustained reduction in body weight along with an early transient decrease in pulmonary capillary wedge pressure without activation of the RAAS. However, over the long term there is a reactive increase in vasopressin levels, with a resultant increased activation of the unprotected V1a receptor in the myocardium and vascular wall. This increase in vasopressin levels in patients receiving tolvaptan has recently been associated with an increase in aldosterone levels.