Meander dynamics has been the focus of river engineering for decades; however, it remains a challenge for researchers to precisely replicate natural evolution processes of meandering channels with numerical models due to the high nonlinearity of the governing equations. The present study puts forward a nonlinear model to simulate the flow pattern and evolution of meandering channels. The proposed meander model adopts the nonlinear hydrodynamic submodel developed by Blanckaert and de Vriend, which accounts for the nonlinear interactions between secondary flow and main flow and therefore has no curvature restriction. With the computational flow field, the evolution process of the channel centerline is simulated using the Bank Erosion and Retreat Model (BERM) developed by Chen and Duan. Verification against two laboratory flume experiments indicates the proposed meander model yields satisfactory agreement with the measured data. For comparison, the same experimental cases are also simulated with the linear version of the hydrodynamic submodel. Calculated results show that the flow pattern and meander evolution process predicted by the nonlinear and the linear models are similar for mildly curved channels, whereas they exhibit different characteristics when channel sinuosity becomes relatively high. It is indicated that the nonlinear interactions between main flow and secondary flow prevent the growth of the secondary flow and induce a more uniform transverse velocity profile in high-sinuosity channels, which slows down the evolution process of meandering channels.