Regional citrate anticoagulation (RCA) is being used more commonly in children for continuous renal replacement therapy. Few reports describe the application of membrane-based therapeutic plasma exchange (mTPE) with RCA in children with liver failure (LF). To explore the application of RCA-mTPE in children with LF. We retrospectively analyzed data from children with LF who underwent RCA-mTPE in the Children's Hospital of Chongqing Medical University's pediatric intensive care unit. We used the total to ionized calcium ratio (T/iCa) > 2.5 as the diagnostic criteria for citrate accumulation (CA). The patients were divided into two groups according to the occureence of CA at the end of RCA-mTPE (CA group: T/iCa > 2.5; NCA group: T/iCa ≤ 2.5). To evaluate the clinical safety and efficacy of RCA-mTPE, the following data from medical records were assessed and compared between groups: clinical characteristics, reasons for LF, RCA-mTPE parameters and duration, laboratory findings, and complications. In total, 92 RCA-mTPE treatments were administered to 21 children with LF over 3.8 ± 0.9 h. The following mean values were determined: blood flow rate (QB) = 2.8 ml/kg/min, 4% sodium citrate dose/blood flow rate ratio (QCi/QB) = 1.1(QCi,ml/kg/h); plasma dose/body weight ratio(QP/BW) = 18.5 (QP, ml/kg/h); 10% calcium gluconate dose/blood flow rate ratio (QCa/QB) = 0.2(QCa, ml/kg/h). The mean concentration of iCa in vitro was 0.38 ± 0.07 mmol/L. Citrate accumulation was recorded after 34 (37%) treatments. Hypocalcemia occurred in 11 (12%) and 7 (7.6%) treatments, during and after mTPE, respectively. Three hypotensive and one convulsive events, related to hypocalcemia, and two clotting events occurred during RCA-mTPE. After RCA-mTPE, the patients' pH, HCO3- and Na+ levels, and T/iCa were significantly increased and the total bilirubin (TB), conjugated bilirubin (DB), prothrombin time (PT), activated partial thromboplastin time (APTT), alanine aminotransferase (ALT), aspartate aminotransferase (AST),and ammonia levels were significantly decreased. The TB, DB, and lactic acid levels, before RCA-mTPE, were significantly higher in the CA group than in the NCA group, but there were no significance between the two groups in QB/BW, QCi/QB, and QP/BW, mTPE duration, and estimated amount of citrate metabolized. Children with LF undergoing RCA-mTPE are at risk of hypocalcemia. With proper protocol adjustment, however, RCA-mTPE can be used safely and effectively in these patients.