The aim of this study was to develop physiologically based pharmacokinetic (PBPK) models capable of simulating cefadroxil concentrations in plasma and tissues in mouse, rat, and human. PBPK models in this study consisted of 14 tissues and 2 blood compartments. They were established using measured tissue to plasma partition coefficient ( ) in mouse and rat, absolute expression levels of hPEPT1 along the entire length of the human intestine, and the transporter kinetic parameters. The PBPK models also assumed that all the tissues were well-stirred compartments with perfusion rate limitations, and the ratio of the concentration in tissue to the unbound concentration in plasma is identical across species. These PBPK models were validated strictly by a series of observed plasma concentration-time profile data. The average fold error (AFE) and absolute average fold error (AAFE) values were all less than 2. The models' rationality and accuracy were further demonstrated by the almost consistent calculated by the PBPK model and noncompartmental method, as well as the good allometric scaling relationship of and . The model suggests that hPEPT1 is the major transporter responsible for the oral absorption of cefadroxil in human, and the plasma concentration-time profiles of cefadroxil were not sensitive to dissolution rate faster than T = 2 h. The cefadroxil PBPK model in human is reliable and can be used to predict concentration-time profile at infected tissue. It may be useful for dose selection and informative decision-making during clinical trials and dosage form design of cefadroxil and provide a reference for the PBPK model establishment of hPEPT1 substrate.