Several reports have demonstrated a close correlation between plasma atrial natriuretic peptide (ANP) concentration and atrial pressure in stable heart diseases. However, few studies have investigated whether plasma ANP concentration is a noninvasive indicator of hemodynamic parameters during the treatment of heart failure. Thus, we have studied the relationship between peripheral plasma ANP concentration and concurrent hemodynamic variables during the treatment of heart failure, and, in order to determine whether secretion of ANP is stimulated in this disease condition, we compared the plasma ANP concentration in the pulmonary artery with that in the peripheral veins. Studies were performed in each of 9 patients with acute heart failure due to myocardial infarction (Group A) or chronic heart failure (Group B), who were matched as closely as possible for treatment, age, sex and cardiac output. In group A, no significant correlation was found between plasma ANP levels and any measured hemodynamic variables. In group B, peripheral plasma ANP concentrations were significantly correlated with left atrial pressure (r = 0.82, p less than 0.01), but not with right atrial pressure (r = 0.56, p greater than 0.05). Furthermore, in group B ANP levels in pulmonary arterial plasma were consistently higher than those in peripheral venous plasma, whereas in group A the opposite was observed in expired cases. These results suggest that measurement of peripheral plasma ANP is a useful noninvasive method for estimating left atrial pressure during the treatment of chronic heart failure. However, plasma ANP concentration may not be a valid means of estimating hemodynamic parameters in acute heart failure due to myocardial infarction. In such cases, the increased secretion of ANP was not obvious, and there may be other factors, in addition to atrial pressure, that regulate cardiac secretion of ANP.