Background: Asymmetric dimethylarginine (ADMA) as a uremia toxin is accumulated in end-stage renal disease (ESRD) patients. Elevated ADMA level has been shown to be predictive of cardiovascular diseases (CVDs) and all-cause mortality in ESRD. Therefore, we investigated the removal of ADMA by different dialysis treatments. Methods: There were 30 each of hemodialysis (HD), hemodiafiltration (HDF), peritoneal dialysis (PD) patients, and healthy volunteers enrolled. The ADMA concentrations in serum, urine, and spent dialysate samples were determined. The urine and spent dialysate volumes were recorded. The ADMA removals by urine and spent dialysate in 1 week were calculated and compared among four groups. It was also analyzed for the correlations between the total removal of ADMA in 1 week and the parameters of age, durance of dialysis, glomerular filtration rate, urine volume, urinal ADMA level, spent dialysate volume, and spent dialysate ADMA level. Results: The serum levels of ADMA in dialysis patients were much higher than in healthy subjects (0.32 +/- 0.09 mu mol/L), and their 1-week total removals of ADMA were much lower than healthy controls (249.21 +/- 57.04 mu mol/week) (p-values all were less than 0.01). Among dialysis groups, serum ADMA levels decreased significantly in PD patients compared with HD or HDF patients (1.38 +/- 0.30 mu mol/L vs. 1.82 +/- 0.38 mu mol/L and 1.63 +/- 0.32 mu mol/L, p < 0.01), and the total removal of ADMA diminished remarkably by turns of PD, HDF, and HD groups (47.79 +/- 8.20 mu mol/week, 31.79 +/- 8.92 mu mol/week, 14.63 +/- 6.53 mu mol/week, respectively, p < 0.01). The total removal of ADMA in 1 week was related directly with the spent dialysate concentrations of ADMA, the spent dialysate volume, and the urine volume. Conclusions: ADMA was mainly removed by dialysate in dialysis patients. Different dialysis models have different clearance capability on plasma ADMA. PD might be more effective on ADMA removal than HD and HDF, with HDF being more effective than HD.