Background and Purpose: The complex pharmacokinetic (PK) properties of monoclonal antibody (mAb) including a nonlinear pharmacokinetics and a significant individual variation in PK process can not be appropriately described by classic PK models, probably derived form a poor understanding of mAbs complex elimination.In our study, a novel PK model based on complex drug elimination was built and evaluated for bevacizumab PK studies in rabbits and beagle dogs.Approach: A new PK model was built based on a comprehensive understanding of mAbs elimination.Subsequently, this model was used to fit the single-dose plasma concentration data of bevacizumab in rabbits and beagle dogs.Moreover, it was also utilized to predict bevacizumabs multiple-dose PK profiles by employing the parameters derived from a fit with the single-dose PK data.All outcomes of model fitting and prediction were compared with those fitted with classic PK models.Key Results: The novel model reasonably described the single PK profiles of bevacizumab in rabbits and beagle dogs, and its fitting capacity was greatly improved compared with compartment models based on the residual sum of squares.The variations existing in the new models parameters could reasonably explain the individual variations of bevacizumabs PK profiles.Moreover, the new model successfully predicted the multiple-dose PK profiles of bevacizumab in beagle dogs.Conclusions and Implications: The novel model reasonably explained the complex elimination of bevacizumab, and exhibited a potential as a useful tool for the PK studies of mAb in practice.