OBJECTIVE To establish whether estimation of creatinine clearance (CrCl) from serum creatinine, gender, age, and weight might reduce the number of 24-hour urine and dialysate collections required to monitor adequacy of delivered dialysis on continuous ambulatory peritoneal dialysis (CAPD). METHODS Retrospective single-center study. METHODS University Hospital. METHODS Creatinine excretion and CrCl were measured in 187 24-hour urine and dialysate collections from 99 CAPD patients (55 male, 44 female). Multiple regression analysis was used to estimate creatinine excretion from age and weight in males and females. CrCl was derived and also calculated using the Cockcroft-Gault and Mitch-Walser formulas. Positive and negative predictive values for indicating adequacy of dialysis were determined. RESULTS Measured and derived CrCl were correlated (males: r = 0.85; females: r = 0.83; p < 0.001), but agreement was poor (95% limits of agreement: males, 26.05 to -25.75 L/wk; females, 37.47 to -19.49 L/wk). Taking the minimum acceptable CrCl as 60 L/week, the respective positive predictive values of the derived, Cockcroft, and Mitch methods in predicting underdialysis were 88%, 100%, and 100% in males and 88%, 88%, and 89% in females. Negative predictive values were 83%, 57%, and 53% in males and 53%, 48%, and 45% in females. CONCLUSIONS A derived CrCl > 60 was not predictive of adequate dialysis. Because the detection of underdialysis is our objective, formal clearance studies should be performed in this group. A derived CrCl < 60 L/wk was predictive of underdialysis in males and females and an increase in dialysis dose without formal clearance measurements could be suggested in these patients. The use of this approach could allow an important reduction in the number of clearance studies required to monitor CAPD adequacy.