Refilling of the vascular space through absorption of interstitial fluid by micro vessels is a crucial mechanism for maintaining hemodynamic stability during hemodialysis (HD) and allowing excess fluid to be removed from body tissues. The rate of vascular refilling depends on the imbalance between the Starling forces acting across the capillary walls as well as on their hydraulic conductivity and total surface area. Various approaches have been proposed to assess the vascular refilling process during HD, including the so-called refilling coefficient (Kr) that describes the rate of vascular refilling per changes in plasma oncotic pressure, assuming that other Starling forces and the flow of lymph remain constant during HD. Several studies have shown that Kr decreases exponentially during HD, which was attributed to a dialysis-induced decrease in the whole-body capillary hydraulic conductivity (LpS). Here, we employ a lumped-parameter mathematical model of the cardiovascular system and water and solute transport between the main body fluid compartments to assess the impact of all Starling forces and the flow of lymph on vascular refilling during HD in order to explain the reasons behind the observed intradialytic decrease in Kr. We simulated several HD sessions in a virtual patient with different blood priming procedures, ultrafiltration rates, session durations, and constant or variable levels of LpS. We show that the intradialytic decrease in Kr is not associated with a possible reduction of LpS but results from the inherent assumption that plasma oncotic pressure is the only variable Starling force during HD, whereas in fact other Starling forces, in particular the oncotic pressure of the interstitial fluid, have an important impact on the transcapillary fluid exchange during HD. We conclude that Kr is not a good marker of LpS and should not be used to guide fluid removal during HD or to assess the fluid status of dialysis patients.