B-type natriuretic peptide (BNP) has been found to be inversely related to peak oxygen consumption (peak VO) in various patient populations. However, in these studies, circulating plasma BNP, i.e. the net effect of release and elimination, rather than cardiac BNP release has been measured. We assessed the relationship between the transcardiac BNP gradient [ΔBNP, i.e. the difference between BNP in coronary sinus (BNP) and arterial (BNP) plasma] and peak VO in healthy subjects with a view to better understanding the regulation of cardiac BNP release in humans.2We studied 10 asymptomatic subjects (age 64±11 years, two females) with preserved left ventricular function (left ventricular ejection fraction 62±5%, averaged early diastolic mitral annular velocity 9±3cm/s) and low BNP (BNP in venous plasma [BNP] <100ng/l). Subjects underwent measurement of BNP and BNP for the calculation of ΔBNP, maximal cardiopulmonary exercise testing, echocardiography and resting and submaximal exercise right heart catheterisation.VThe median (range) BNP, BNP, BNP, and ΔBNP were 62 (14, 82), 60 (13, 79), 110 (25, 157), and 44 (1, 103) ng/l. The median peak VO during cardiopulmonary exercise testing was 21.5 (18, 54) ml/min/kg. There was an inverse correlation between higher ΔBNP and lower peak VO (r=-0.84; p=0.002) and oxygen pulse (r=-0.64, p=0.049). There was a trend towards an inverse correlation between ΔBNP and the exercise arteriovenous oxygen content difference (r=-0.58; p=0.08).VIn healthy humans, there is an inverse association between myocardial BNP release and peak VO, which may be due to cardiac and non-cardiac mechanisms.2