This paper proposes an energy flow model and an optimal energy management strategy based on mean values for maximizing driving mileage of a fuel cell bus (FCB), which is powered by a polymer electrolyte membrane (PEM) fuel cell system and a lithium battery. Firstly, an energy flow model describing the relations between vehicle performance and power flow parameters is quantitatively established. An optimization problem for maximizing driving mileage on a predetermined route is defined, and an analytical solution with clear physical meanings is derived. Next, a practical real-time supervisory Energy Management strategy basing on Mean Values (EMMV) is proposed. The strategy, which doesn't require a priori knowledge of the driving trip, is then compared with several well-known strategies, e.g. charge depleting and charge-sustaining (CDCS), Blended, dynamic programming (DP), and Pontryagin's Minimum Principle (PMP). Simulation results show that, the proposed strategy achieves a near-optimal effect, and converges after one driving cycle on a predetermined bus route. Finally, on-road testing is carried out. The proposed strategy achieves an average endurance mileage on a real bus route of 162 km with a usable battery state of charge (SOC) of 90% and 20 kg hydrogen gas for a fully loaded fuel cell city bus. Copyright (C) 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.