The present paper develops a real-time clutch transition strategy for a parallel hybrid electric vehicle (HEV) in order to achieve quick and smooth clutch transition engagements between pure electrical driving and hybrid driving. Model predictive control (MPC) has been used for this model and tested with different control horizons and weighting factors to verify the ability of MPC to control the vehicle speeds for the clutch engagement. Some modified MPC algorithms with softened constraints and with output regions have been also studied to improve the robustness and the ability of this controller. Comprehensive simulations for the HEV have been conducted in MATLAB and Simulink. Results show that the system can provide real-time optimal control actions subject to input and output constraints for real-time clutch transition engagement with high driving comfort. The system can be implemented in electronic control units and applied for real HEVs. © 2012 IMechE.