OBJECTIVES Automated peritoneal dialysis (APD) has been increasingly used since payment scheme was expanded to cover APD machine in addition to the continuous ambulatory peritoneal dialysis (CAPD) in the benefit package of National Health Insurance Program in May of 2008. This study aims to compare the health outcome between patients who used APD and CAPD. METHODS The including criteria were patients treated by APD or CAPD identified in National Health Insurance Research Database (NHIRD) during 2001-2010. The excluding criteria were patients who were treated by hemodialysis for more than 3 months before PD, or younger than 18 years old, or received kidney transplant before. The CAPD patients and APD patients were identified and matched according to their propensity score predicted by age, gender, comorbid conditions, Charlson Comorbid Index, medication history, and premiums wages in the year of treatment initiation. There were 2,287 APD and 2,287 CAPD patients entered the final analysis. The Kaplan-Meier curve and the Cox proportional hazard regression were performed to examine the differences in mortality rate, technique failure rate and incident rate of peritonitis between APD and CAPD. RESULTS There are significant differences in mortality rate (0.76% vs. 0.65% per patient month, p=0.02), in technique failure rate (1.07% vs. 0.87% per patient month, p<0.001), and in incidence rate of peritonitis (1.94% vs. 2.14% per patient month, p<0.01) between APD and CAPD cases. When compared with CAPD, patients treated by APD posed higher risk in death and technique failure but no difference in the incidence of the 1^stperitonitis with hazard ratios of 1.18 (CIs: 1.03-1.34), 1.22 (CIs: 1.09-1.37) and 0.98 (CIs: 0.88-1.10), respectively. CONCLUSIONS The APD patients seemed to have higher mortality rate and technique failure rate than CAPD patients, however, APD patients had lower incident rate of peritonitis than CAPD patients in Taiwan.