We examined the effect of renal dysfunction on B-natriuretic peptide (BNP), N-terminal (NT)-proBNP, and their molar ratio at varying severities of cardiac function in 94 Thai patients with chest pain (52 men; 32 women), also measuring creatinine and left ventricular ejection fraction (LVEF). Renal function was classified into 5 stages by estimated glomerular filtration rate. The molar NT-proBNP/BNP ratio was calculated. Cardiac status was classified by LVEF (normal, >50%; moderate, 35%-50%; severe, <35%). BNP, NT-proBNP, and their ratio corresponded to renal disease stage exponential (0.51, 1.05, and 0.54, respectively; correlation coefficients, >or=0.95). BNP and the ratio are affected less than NT-proBNP by renal dysfunction, starting in stage III; NT-proBNP expresses effects starting in stage II. NT-proBNP is more sensitive than BNP to renal disease stage. For log of geometric means vs stage of renal disease, the BNP slopes and correlation coefficients vary considerably (slopes, 0.036-0.531; r(2), 0.017-0.99). The NT-proBNP slopes and regression coefficients vary considerably (slopes, 0.18-0.71; r(2), 0.33-0.99). For the ratio, the slopes show low variation (0.148-0.337), r(2) greater than 0.96, women differing from men (P = .012). The effect of renal disease differs by gender. BNP and NT-proBNP increase by stage III for women but not for men. One must consider renal function, gender, and LVEF when using BNP or NT-proBNP as cardiac biomarkers. The ratio of the 2 peptides is the most consistent marker across LVEFs.