B-type or brain natriuretic peptide (BNP) is a novel natriuretic peptide secreted from the heart that forms a peptide family with A-type or atrial natriuretic peptide (ANP), and its plasma level has been shown to be increased in patients with congestive heart failure. This study was designed to examine the sources and mechanisms of the secretion of BNP in comparison with those of ANP in control subjects and in patients with heart failure. We measured the plasma levels of BNP as well as ANP in 16 patients with dilated cardiomyopathy (11 men and 5 women; mean age, 59 years) and 18 control subjects (9 men and 9 women; mean age, 54 years) by sampling blood from the femoral vein, the aortic root, the anterior interventricular vein (AIV), and the coronary sinus using the newly developed immunoradiometric assay systems. In the control subjects, there was no significant difference in the plasma ANP level between the aortic root and the AIV (24.0 +/- 5.2 pg/mL versus 32.2 +/- 17.0 pg/mL), but there was a highly significant step-up of the level between the AIV and the coronary sinus (32.2 +/- 17.0 pg/mL versus 371.4 +/- 111.1 pg/mL, P < .001). In contrast, there was a significant step-up of the plasma BNP level between the aortic root and the AIV (8.6 +/- 6.4 pg/mL versus 19.0 +/- 11.5 pg/mL, P < .01) but not between the AIV and the coronary sinus (19.0 +/- 11.5 pg/mL versus 28.8 +/- 14.0 pg/mL). On the other hand, in patients with dilated cardiomyopathy, there was a significant step-up in the plasma ANP level between the aortic root and the AIV (280.6 +/- 183.7 pg/mL versus 612.3 +/- 431.6 pg/mL, P < .01) and between the AIV and the coronary sinus (612.3 +/- 431.6 pg/mL versus 1229.0 +/- 772.7 pg/mL, P < .01). There was a significant step-up in the plasma BNP level between the aortic root and the AIV (268.4 +/- 293.2 pg/mL versus 511.6 +/- 458.1 pg/mL, P < .01) but not between the AIV and the coronary sinus (511.6 +/- 458.1 pg/mL versus 529.7 +/- 455.3 pg/mL) in patients with dilated cardiomyopathy. The arteriovenous difference at the AIV of the plasma level of BNP had a significant positive correlation with left ventricular end-systolic volume index (r = 0.859, P < .001) and a significant negative correlation with left ventricular ejection fraction (r = -.735, P < .001). We conclude that (1) BNP is secreted mainly from the left ventricle in normal adult humans as well as in patients with left ventricular dysfunction, whereas ANP is secreted from atria in normal adult humans and also from the left ventricle in patients with left ventricular dysfunction; (2) secretion of BNP as well as ANP from the left ventricle increases in proportion to the severity of the left ventricular dysfunction, suggesting that the secretions of ANP and BNP from the left ventricle are regulated mainly by wall tension of the left ventricle; and (3) the peripheral plasma levels of ANP and BNP reflect the secretion rate of these hormones from the left ventricle and may be used as a marker of the degree of left ventricular dysfunction in patients with left ventricular dysfunction.