The important neuroendocrine systems involved in heart failure are reviewed with special emphasis on their possible role in pathophysiology and their relation to prognostic and diagnostic information. Plasma levels of noradrenaline (NA), renin, vasopressin, endothelin-1, atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP) and tumour necrosis factor-alpha (TNF-alpha) are all elevated in heart failure. Activity of the sympathetic nervous system as reflected by NA is correlated to mortality and seems to possess independent prognostic information. Several studies have now documented the beneficial effect of beta-blockade in chronic heart failure (CHF). Renin seems to be a poor prognostic marker in CHF possibly because of the interference with diuretic treatment, angiotensin converting enzyme (ACE)-inhibitors and angiotensin II antagonist, and probably also because of the significance of tissue renin-angiotensin system (RAS), poorly reflected by plasma renin. On the other hand, several large-scale trials with ACE-inhibitors and angiotensin II antagonists have demonstrated reduced mortality and morbidity in CHF. Plasma vasopressin does not seem to possess prognostic information but testing of non-peptide antagonists is ongoing. Endothelin-1 seems to have independent prognostic information and endothelin receptor antagonists may represent a therapeutic possibility. The natriuretic peptides ANP and BNP are correlated to prognosis and possess independent information. Brain natriuretic peptide and N-terminal ANP seem to increase early, i.e. in asymptomatic heart failure. Plasma BNP being more stable than ANP is therefore a promising measure of left ventricular dysfunction. Increase in ANP and BNP, potentially beneficial, may be achieved by administration of neutral endopeptidase inhibitors, at present an unsettled therapeutic possibility. Several cytokines are increased in heart failure and especially TNF-alpha has drawn attention. Experimental studies suggest that TNF-alpha is important in the pathophysiology of heart failure and preliminary studies indicate that inhibition of TNF-alpha seems to be a possible therapeutic approach. Thus, neuroendocrine markers seem to (i) have a role in diagnosis and classification of heart failure, (ii) be useful in providing a 'neuroendocrine profile' which enlightens different aspects of heart failure, and therefore (iii) in the future probably will be valuable in the choice of medical treatment of the individual patient. In addition to beta-blockers, ACE-inhibitors and angiotensin II antagonists several new drugs based on neuroendocrine modification are on their way and might become important in the future.