Hemodialysis (HD) is a method used to remove biogenic substances or blood components that cause disease and some drugs used by patients to treat their diseases. Therefore, dosing schedule must be planned according to HD clearance (CL_HD) when medical treatment is provided to patients receiving HD. We aimed to clarify the physical properties (eg, octanol‐water partition coefficient and molecular electronegativity) or pharmacokinetic parameters (eg, volume of distribution) of compounds affecting CL_HD and to construct a mathematical model to predict CL_HD. The analysis covered individual CL_HD data for nine compounds from the literature. The molecular descriptors which are physical properties or pharmacokinetic parameters were calculated using the structural formula of each compound, and searched for factors related to CL_HD among the calculated 148 molecular descriptors. Nonlinear mixed‐effects model analysis with CL_HD as objective variable and molecular descriptors as explanatory variable was conducted to examine the factor affecting CL_HD and develop a model for predicting CL_HD. The logarithm of the brain/blood partition coefficient was detected as a factor affecting CL_HD. The predictive accuracy of CL_HD using the constructed mathematical model with the logarithm of the brain/blood partition coefficient as explanatory variable was adequate.