Abstract Background The prognostic utility of N-terminal pro-B-type (NT-proBNP) in patients with heart failure and chronic kidney disease (CKD) is incompletely understood since elevations in NT-proBNP could be related either to decreased clearance by the kidney or to increased severity of structural heart disease. Purpose We sought to assess the association of NT-proBNP with cardiovascular and mortality outcomes in patients with heart failure and mildly reduced or preserved ejection fraction, stratified by baseline kidney function. Methods We conducted a pooled analysis of individual participants with NT-proBNP and estimated glomerular filtration rate (eGFR) measured at baseline in the I-PRESERVE, TOPCAT (Americas region), PARAGON and DELIVER trials. We evaluated the relationship between NT-proBNP and kidney function using piecewise linear regression. Using multivariable Cox and Poisson regression models, we assessed the association of NT-proBNP with clinical outcomes across different levels of eGFR (≥60, 45-<60 and <45 mL/min/1.73m2). The primary outcome was hospitalisation for heart failure or cardiovascular death. Results Among 14,831 participants (mean age 72.1 years, 50.3% female, mean eGFR 63.3 mL/min/1.73m2 and median NT-proBNP 840 pg/mL) followed for a median 33.5 months, there were 3,092 primary outcomes, 2,184 heart failure hospitalisations, 1,465 cardiovascular and 1,049 non-cardiovascular deaths. Participants in lower eGFR categories were more likely to be older, female, and have atrial fibrillation and diabetes (all p<0.001). NT-proBNP levels increased by 9%, 8%, and 23% per 10 mL/min/1.73m2 lower eGFR in patients with baseline eGFR ≥60, 45-60, and <45 mL/min/1.73m2, respectively (P for non-linearity <0.001). For any given NT-proBNP concentration, participants with eGFR <45 mL/min/1.73m2 were at highest risk (Figure 1). Each doubling in NT-proBNP was associated with a 37% relative increase in the primary outcome (HR 1.37, 95% CI 1.34-1.41), consistent across different eGFR categories (P-interaction=0.42). Association between NT-proBNP and other clinical outcomes were also consistent across different levels of kidney function (all P-interaction>0.3; Figure 1). Equivalent risk for the primary outcome was apparent at NT-proBNP levels approximately 2.5- to 3.5-fold lower in patients eGFR <45 mL/min/1.73m2, compared to those with eGFR ≥60 mL/min/1.73m2. (Figure 2). Conclusion NTproBNP is a potent predictor of risk in patients with heart failure with mildly reduced or preserved ejection fraction across the spectrum of kidney function. However, in patients with reduced kidney function, lower NT-proBNP levels confer absolute risk of cardiovascular outcomes similar to substantially higher NT-proBNP levels in patients with normal kidney function. Accordingly, interpretation of NTproBNP levels should vary according to kidney function, particularly in those with eGFR <45 mL/min/1.73m2.