Venous catheters have become an indispensable form of hemodialysis access. We evaluated catheter performance as temporary and long-term access in children with end-stage renal disease (ESRD). We assessed the survival rates and causes of catheter failure in 78 catheters used for hemodialysis access in 23 pediatric patients (aged 10 months to 22 years) with ESRD over a 5-year period. Median survival was 31 days for 56 uncuffed catheters. One- and 2-month actuarial survival was 69% and 48%, respectively. Reasons for removal were: elective (39%), kinking (36%), trauma (11%), infection (7%), and other (5%). Smaller catheters (7 or 9 French) were more likely to be removed for kinking (P = 0.003). One-year actuarial survival for 22 cuffed catheters was 27%. Cuffed catheters were removed due to: infection (36%), kinking (14%), elective (9%), trauma (9%) and other (9%). Twelve catheters were removed for infection. Infection rates leading to removal were 0.58 and 0.71 per patient-year for uncuffed and cuffed catheters, respectively. Staphylococcus species were cultured most commonly. We conclude that uncuffed catheters function well for short-term hemodialysis access of up to 2 months' duration and cuffed catheters are successful for long-term access in children and adolescents with ESRD.