This study examines the frequency of discrepancy between Kt/V urea and creatinine clearance (Ccr) measurements in patients on peritoneal dialysis (PD) and the reasons for this discrepancy. METHODS Nonrandomized, retrospective data analysis. METHODS Single PD unit of a university teaching hospital. METHODS All adult patients receiving PD at our center from January 1995 to December 1996. METHODS Actual (a) and desired (d) body weight (BW) were used to calculate urea volume of distribution (V) and body surface area (BSA). Patients were divided into four groups based upon their total small solute clearances (Kt/V and Ccr, normalized by actual weight) and three additional groups based upon actual/desired (a/d) body weight ratio. An additional analysis was performed for the subset of anuric patients. Data collected for all patients included the following: total Kt, total Ccr, 4-hour dialysate/ plasma (D/P) creatinine, serum albumin concentration, duration of PD, actual body weight, age, and height. RESULTS Twenty-three percent of the clearance measurements in our study were discrepant, defined as having values for either Kt/V or Ccr (but not both) above the accepted targets of Kt/V > or = 2.0/wk and Ccr > or = 60 L/wk/ 1.73 m2. Patients with both values above target are more likely to have higher residual renal function. Patients who are significantly less than BWd and patients on PD for a longer time are more likely to have adequate Kt/V but not Ccr. Furthermore, patients who are less than 90% or greater than 110% of BWd have markedly different values for Kt/V and Ccr when BWa versus BWd values are used. CONCLUSIONS Kt/V and Ccr values are frequently discrepant; a number of factors affect these two measurements to varying degrees, including weight, degree of residual renal function, and duration of PD.