The importance of mathematical modeling in dialysis arose recently as a consequence of the achievement of individualized treatment therapies. Two distinct models were used in our center, namely the urea kinetic model (UK) and the direct dialysis quantification (ddq), for the estimation of the urea distribution volume (V) and the urea generation rate (G). It turned out that, for the same patient, they provided us with different results. The basic hypotheses on which the two models rely are poorly described in the literature. The aim of this paper is an attempt to fill this gap and to show that the discrepancy observed between the two models is not surprising at all. The study also shows that from a theoretical viewpoint, ddq is more appropriate than UK for the estimation of V and G. The question of the intrinsic quality of ddq is not discussed here.