The present single-center cohort study was based on a clinical intensive care unit database containing data on 1128 consecutive children undergoing their first operation for congenital heart disease between 1993 and 2002 at Aarhus University Hospital, Skejby, Denmark. A total of 130 (11.5%) children developed postoperative acute renal failure (ARF) managed with peritoneal dialysis (PD). Logistic regression analysis was used to examine risk factors for complications related to PD and to compare mortality between ARF and non-ARF patients controlling for potential confounding factors. A total of 43 complications related to PD were registered in 27 (20.8%) patients. Major complications were seen in eight (6.2%) patients, and only two (1.5%) patients were switched to hemodialysis after peritonitis and hemicolectomy due to bowel perforation. The main risk factors for complications to PD were duration of PD, high RACHS-1 score (Risk Adjusted Classification for Congenital Heart Surgery), and hyperkalemia at initiation of PD. Overall, in-hospital mortality was 6.8% (76/1128). Mortality of ARF patients was 20.0% compared to 5.0% among non-ARF patients (adjusted odds ratio=1.91, 95% confidence interval=1.10-3.36). After stratification, ARF was strongly associated with increased mortality in the subgroups of patients with the lowest overall risk of dying (age> or =1 year, body weight> or =5 kg, RACHS-1 score <3, and no preoperative cyanosis). For patients at high risk of dying (age <1 year, body weight <5 kg, RACHS-1 score> or =3, cardiopulmonary bypass time> or =60 min, and preoperative cyanosis), the association between ARF and mortality was substantially weaker. In conclusion, postoperative ARF was associated with increased mortality in children operated for congenital heart disease. Major complications to PD were few, and our data strongly support that PD is a simple, safe, feasible, and robust dialysis modality for the management of ARF in children.