Hybrid　electric　vehicle　is　considered　as　the　most　promising　solution　to　energy　crisis　and　urban　air　pollution.　However,　as　it　has　at　least　two　sets　of　propulsion　systems,　its　configuration　is　complex.　To　get　better　fuel　economy　and　vehicle　performance,　the　design　and　sizing　of　its　powertrain　components　are　important.　In　this　paper,　an　HEV　optimal　sizing　method　combining　optimization　algorithm　and　HEV　simulation　tool　is　introduced,　and　then　a　real-coded,　adaptive　based　hybrid　genetic　algorithm　is　developed　and　applied　to　the　optimal　sizing　of　a　series　hybrid　electric　vehicle.　ADVISOR2002　is　used　as　the　vehicle　simulator.　The　results　have　proved　the　validity　of　the　optimal　sizing　methodology　and　the　efficiency　of　the　hybrid　genetic　algorithm.　Based　on　the　results,　some　improvements　are　proposed　on　the　vehicle　studied.