The paper presents the development of a morphological model and its application to experimental model rivers.The model takes into account the key processes of channel migration,including bed deformation,bank failure and wetting and drying.Secondary flows in bends play an important role in lateral sediment transport,which further affects channel migration.A new formula has been derived to predict the near-bed secondary flow speed,in which the magnitude of the speed is linked to the lateral water level gradient.Since only non-cohesive sediment is considered in the current study,the bank failure is modelled based on the concept of submerged angle of repose.The wetting and drying process is modelled using an existing method.Comparisons between the numerical model predictions and experimental observations for various discharges have been made.It is found that the model predicted channel planform and cross-sectional shapes agree generally well with the laboratory observations.A scenario analysis is also carried out to investigate the impact of secondary flow on the channel migration process.It shows that if the effect of secondary flow is ignored,the channel size in the lateral direction will be seriously underestimated. The paper presents the development of a morphological model and its application to experimental model rivers. The model takes into account the key processes of channel migration, including bed deformation, bank failure and wetting and drying. Secondary flows in bends play an important role in lateral sediment transport, which further affects channel migration. A new formula has been derived to predict the near-bed secondary flow speed, in which the magnitude of the speed is linked to the lateral water level gradient .ince only non-cohesive sediment is considered in the current study, the bank failure is modelled based on the concept of submerged angle of repose. The wetting and drying process is modelled using an existing method. Comparisons between the numerical model predictions and experimental observations for various discharges have been made. It is found that the model predicted channel planform and cross-sectional shapes agree generally well with the laboratory observations. A scenario anal shows that the effect of secondary flow is ignored, the channel size in the lateral direction will be able to be underestimated.