Controlled clinical trials performed in more than 13 000 patients have, to date, consistently shown the beneficial effects of long term β-adrenoceptor antagonist (β-blocker) therapy in patients with chronic heart failure. It is not clear whether this represents a class effect or whether it is specific only to some agents. Beneficial effects on the prognosis of patients with mild to moderate heart failure have been shown with metoprolol, bisoprolol, and carvedilol. These β-blockers, however, differ in their pharmacologic characteristics. Metoprolol and bisoprolol are selective for β1-adrenergic receptors and are devoid of ancillary properties. Carvedilol, at a dosage of 50 mg/day, blocks all β1-, β2-, and α1-adrenergic receptors, and it has associated antiproliferative and antioxidant activities. These differences cause a varied acute hemodynamic response, with a reduction in cardiac output and a tendency toward a rise in pulmonary wedge pressure with selective agents and no change in cardiac output and a slight decrease in pulmonary pressures with carvedilol. Accordingly, when the therapy is started, the most frequent adverse effects are worsening heart failure with metoprolol and bisoprolol, and hypotension and dizziness with carvedilol. It remains controversial whether these differences also influence the long term effects of therapy. Carvedilol may provide a more comprehensive blockade of the cardiac adrenergic drive than selective β-blockers because it does not upregulate β1-adrenergic receptors, blocks all adrenergic receptors and decreases cardiac norepinephrine release. These properties may lead to a larger increase in left ventricular function and a lack of improvement in maximal exercise capacity with carvedilol, compared with selective β-blockers. It is, however, unclear whether these differences also influence patient outcome. The long term effects of different β-blockers on prognosis are currently being compared in the Carvedilol or Metoprolol European Trial (COMET) in which more than 3000 patients with chronic heart failure have been randomized in a 1 : 1 ratio to receive metoprolol or carvedilol. © Adis International Limited. All rights reserved.