Plug-in hybrid electric vehicles (PHEVs) with post-transmission parallel configuration attracted considerable attention due to their capacity to operate in either electric vehicle (EV) mode or hybrid electric vehicle (HEV) mode. Meanwhile, the added flexibility and multiple operation modes add additional challenges to vehicle control with acceptable drivability, particularly during the mode transition from the EV and HEV, since proper control is needed for the internal combustion engine (ICE), motor and coupling device to achieve smooth and fast transition, under various vehicle operation constraints such as mode-transition duration, vehicle acceleration fluctuation and friction loss of the dry clutch. In addition, the engagement of dry clutch features torque discontinuity due to slip-stick phenomenon and the dynamic behavior of the ICE further increases the nonlinearity of the powertrain system. This research introduces a method for identifying the theoretically optimal drivability during mode transition and feasible control schemes to effectively coordinate different powertrain components and achieve desirable drivability without violating vehicle operation constraints. Firstly, a post-transmission parallel PHEV is modeled using MATLAB Simulink and US DOE-ANL's AUTONOMIE, and partially verified using a research prototype PHEV-EcoCAR, developed on the GM Chevrolet Malibu platform at University of Victoria. Secondly, the problem of optimal control for drivability is formulated, considering the vehicle operation constraints. Dynamic Programming (DP) is applied to the Simulink powertrain model to identify the optimal control solutions. Simulation results for the studied vehicle are presented and analyzed to illustrate the new approach and its advantages.