It has been claimed that computed urea kinetic (UK) modelling in haemodialysed patients for the estimation of protein intake and of the relation between total dialyser urea clearance and distribution volume (Kt/V) leads to an overestimation of protein catabolic rate (PCR). In the present study three different methods of kinetic modelling for the determination of PCR and Kt/V are compared in 15 patients. The first method (MI) is the direct quantification method based on the collection of all urea eliminated from the body. The two other methods are based on an iterative computed calculation. The second method (MII) is the urea kinetic modelling method as described by Sargent. Dialyser clearances were measured directly and not estimated by theoretical extrapolation. The third method described here (MIII), is based on the indirect calculation of urea distribution volume (Vw) according to Watson and of dialyser clearances from this Vw and from pre- and post-dialysis urea concentrations. All three methods result in PCRs that are not significantly different (MI: 1.04 +/- 0.29; MII: 1.07 +/- 0.28; MIII: 1.05 +/- 0.24 mg/kg BW per 24 h; p greater than 0.05). When the results are correlated, the following results are obtained: MI vs MII: r = 0.76, p less than 0.001; MI vs MIII: r = 0.78, p less than 0.001; MII vs MIII: r = 0.90, p less than 0.001. For Kt/V virtually identical results were obtained for each of the methods under study. In conclusion, all methods under study seem equally reliable in determining mean PCR.(ABSTRACT TRUNCATED AT 250 WORDS)