This study was performed to determine compliance with prescribed exchanges in continuous ambulatory peritoneal dialysis (CAPD) and continuous cycler peritoneal dialysis (CCPD) patients via a home supply inventory, to evaluate numerous other factors against this standard, and to compare compliant and noncompliant patient characteristics and outcomes. It was an open enrollment of patients with prospective follow-up of 49 CAPD and CCPD adult patients at a university dialysis center. Two home visits for dialysis solution inventories were conducted 1 to 3 months apart. The number of exchanges actually performed (based on the inventory and deliveries) divided by the number of exchanges prescribed determined the percent compliance. Patient self-evaluation of compliance, attitudes about compliance and medical care, Derogatis Affects Balance Sheet (DABS) (a validated tool of affects balance), staff evaluations, patient demographics, hospitalizations, dialysis adequacy (measured to predicted creatinine ratio) and patient outcomes were evaluated against compliance with prescribed exchanges. Thirty-five percent of the patients were found to be noncompliant with prescribed exchanges based on the supply inventory, performing only 74% of exchanges. Compliant and noncompliant patients were not different by age, race, gender, peritoneal dialysis (PD) time, number of comorbid conditions, nor incidence of diabetes. The affects balance score from the DABS test was a poor predictor of compliance, as were staff evaluations. Delivered dialysis, adjusted for percentage of compliance, was significantly lower in noncompliant than in compliant patients by weekly Kt/V (2.2 v 1.74, P < 0.003), Ccreatinine L/week/1.73m2 (69 v 58, P = 0.05), as was normalized protein nitrogen appearance rates (nPNA), g/kg/day (0.89 v 0.70, P = 0.001). Technique survival was significantly lower (P < 0.05) in noncompliant patients. Twenty-nine percent of the noncompliant patients transferred to hemodialysis for uremia compared with 6% of compliant patients, P = .04, with a mean follow-up time of 1 year per patient. Patient survival did not differ in the two groups. Peritonitis rates were higher in noncompliant than compliant patients (1.03/yr v 0.66/yr, P < 0.03), as were days hospitalized (908 per 100 patient-years v 1,016 per 100 patient-years, P < 0.04). Noncompliance with PD exchanges is significantly common in CAPD and CCPD patients. It occurs in one third of the patients at our center and contributes significantly to technique failure, inadequate dialysis, and an increased risk of both peritonitis and hospitalized days. Home visits to conduct supply inventories provide an excellent method of monitoring compliance.