Despite recent advances in peritoneal dialysis (PD) systems, peritonitis is a significant clinical problem in patients on PD. Risk factors for peritonitis are identifiable and modifiable and require focused intervention. During a baseline period in 1998, we observed consistent differences in peritonitis rates among patients using various PD connection systems. In January 1999, motivated by a need to reduce peritonitis, we initiated a multifaceted continuous quality initiative (CQI) program that included retraining all current patients and all new patients 6 months after initiation and then annually; changing from plastic to titanium adapters between the catheter and the transfer set; and using equipment from a single PD manufacturer for all new patients and for current patients with high peritonitis rates. Furthermore, all patients using HomeChoice cyclers (Baxter Healthcare Corporation, McGaw Park, IL, U.S.A.) were taught to use the Compact Exchange Device II to avoid contamination when spiking solution bags. Peritonitis rates improved from 1 episode per 7.5 patient-months (over 512 patient-months) in 1998 to 1 episode per 36.5 patient-months (over 292 patient-months) as of September 2002. Further analysis also showed a significant difference in peritonitis rates between equipment produced by various manufacturers. There was a statistically significant difference in peritonitis for automated peritoneal dialysis systems. Patients using the Freedom Cycler PD+ (Fresenius Medical Care, Frankfurt, Germany) had an average peritonitis rate of 1 episode per 6.9 patient-months as compared with patients using the HomeChoice cycler (Baxter Healthcare), who experienced 1 episode of peritonitis per 23.9 patient-months on average (p < 0.0001). For continuous ambulatory peritoneal dialysis patients using UltraBag (Baxter Healthcare), the peritonitis rate was 1 episode per 26 patient-months as compared with the Premier Double Bag (Fresenius Medical Care), for which a peritonitis rate of 1 episode per 6.3 patient-months was seen (p < 0.0001). Comparison of the UltraBag (1 episode per 26.0 patient-months) with the Disposable Freedom Set, a single-bag "Y" system (Fresenius Medical Care; 1 episode per 7.5 patient-months) yielded similar results (p < 0.0001). We conclude that ongoing CQI efforts can significantly reduce peritonitis rates. Our efforts included aggressive patient retraining, protocol changes, use of a titanium adapter between the catheter and the transfer set, and careful choice of connectology systems (possible supplier-dependent effect).