In patients with heart failure, increased wall stretch due to volume and pressure overload leads to an increase in circulating natriuretic peptides (ANP and BNP and their N-terminal fragments NT-proANP and NT-proBNP). Plasma BNP levels commonly considered normal (< 20 pg/ml) are influenced by age, sex, and also by genetic factors. ANP and BNP are synthesized and released by atrial and ventricular myocytes (Figure 1). In subjects with acute dyspnea, a BNP plasma concentration of 100 pg/ml has been established as a cutoff value for the diagnosis of heart failure yielding a very high negative predictive value coupled with an acceptable positive predictive value (Figure 3). However, recent evidence suggests that much more subtle elevations of plasma BNP may also indicate an increased long-term risk of cardiovascular events and death (Figure 2). In acute heart failure, natriuretic peptides correlate with ventricular pressure and volume overload, as well as with NYHA functional class. They can, however, not reliably discriminate between heart failure due to reduced ejection fraction and heart failure with preserved systolic function (Figure 4). Thus, elevated BNP or NT-proBNP levels always demand further clarification of heart failure etiology using echocardiography as the method of choice. As indicated by the algorithm for a BNP-based differential diagnosis of acute heart failure symptoms (Figure 5), a variety of noncardiac causes may also lead to moderate elevations of the markers (BNP plasma concentrations of 100-400 pg/ml). In addition, normal marker levels may be observed in > 20% of patients with long-term stable heart failure. Thus, increased plasma concentrations of natriuretic peptides are not strictly specific for heart diseases and also lack sensitivity in the chronic compensated state. Diuretics, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor antagonists, and spironolactone have been shown to decrease BNP and NT-proBNP in parallel with clinical and hemodynamic improvement. In patients hospitalized for decompensated heart failure, predischarge plasma BNP levels reflect the risk of future death and rehospitalization (Figure 6). Although adjusting heart failure treatment to reduce plasma NT-proBNP levels may improve outcome, a general recommendation for monitoring drug therapy using this marker should not be derived from this observation. General practitioners may, in the future, use BNP or NT-proBNP as a rule-out test for heart failure and preselect patients for further diagnostic work-up on the basis of an elevated plasma level. Within the framework of the German network for heart failure the multicentric "Handheld-BNP Study" will clarify, whether echocardiography using low-price simple handcarried devices could be used as an alternative or, more likely, as a complementary diagnostic tool to further improve heart failure diagnosis in primary care.