An online power management strategy (PMS) for plug-in hybrid electric vehicles is presented in this paper. This PMS with the capability to reduce both the fuel consumption and the emission, has simple mathematics and excludes a priori knowledge of the driving cycle. The only required information is the driving duration that can be estimated by the driver or by the vehicle information systems, so the proposed method can be easily implemented. Furthermore, an adaptive form of this PMS is presented and its performance is compared with other strategies. Using the online adaptive PMS method, the incoming driving cycle condition is predicted by the vehicle past conditions. In this paper, the engine fuel characteristics are linearized to several zones. At any instant, one of these zones is selected for the engine operation. In each zone, an optimal cost function is minimized for the fuel consumption and the emission reduction. Moreover, different cost functions are defined and used on various engines. Finally, the proposed PMS is simulated in the ADVISOR environment and compared with conventional method.