Background: Brain natriuretic peptide and its derivative peptide NTproBNP are utilized to exclude cardiac diseases, and predicting risk of mortality in dialysis patients. Our aim was to evaluate both elimination of NTproBNP through dialysate and a possible relationship between plasma and/or dialysate NTproBNP level and membrane transport status of peritoneal dialysis patients. Methods: 57 plasma (P) and dialysate (D) samples of 44 peritoneal dialysis (PD) patients were analyzed for NTproBNP. Modified peritoneal equilibration test (PET) results and other variables were obtained from the charts. Results: Median (IQR) NTproBNP concentrations (pg/mL x 1,000) in P and D were 3.3 (1 - 13) and 0.5 (0.2 - 3.6), respectively. There was a linear correlation between P-NTproBNP and D-NTproBNP (r = 0.928, p = 0.0001; regression equation was y = 0.897(*)x -0.28). Mean P/D-NTproBNP ratio was 5.5 +/- 0.5. Median P and D-NTproBNP levels by the membrane transport status were aligned as high (H) > high average (HA) > low average (LA), and the difference between H and LA was statistically significant (p < 0.001). Mean arterial pressure (MAP), residual Kt/V and dialysate/plasma ratio of crearinine (D/P Cr) were significant predictors of D-NTproBNP; whereas only MAP and residual Kt/V were significant predictors of P-NTproBNP in multiple regression analysis. Both P-and D-NTproBNP have significant and similar size of correlations with MAP, albumin, D/P Cr ratio, and Na. Conclusions: D-NTproBNP level is similar to 1/5 of P-NTproBNP, and the issue of relationship between membrane transport status and natriuretic peptide levels needs more work.