Nutrition has been shown to predict clinical outcome in continuous ambulatory peritoneal dialysis (CAPD) patients. However, despite the positive relationship between KT/V (urea) and the normalised protein catabolic rate, the ability of urea kinetic modelling to predict clinical outcome or nutrition remains inconclusive. We have evaluated the relationship between nutrition and achieved dialysis in a cross-sectional study of 147 stable CAPD patients on dialysis for a mean period of 22 months. Protein-energy malnutrition was present in 22-32% of the study population. 39 and 41% of the population failed to achieve suggested adequacy targets of 50 litres/week for total creatinine clearance and a weekly KT/V(urea) of 1.7, respectively. Severely malnourished patients had significantly greater normalised clearance and adequacy values than well-nourished patients. Intrinsic actual peritoneal clearance bore no relation to patient size. The subsequent normalisation of this value by a component of patient mass results in a mathematical bias against well-nourished or obese patients. This may explain the failure of such adequacy values to reflect outcome and argues against accepting such values as measures of dialysis well-being.