Battery sizing has significant importance for the performance of hybrid electric vehicles (HEVs). Although several research has been done over the years for the battery sizing, no research has focused on battery system efficiency which affects fuel economy. This paper has investigated battery system efficiencies of different optimum battery sizes which were optimised using two design optimisation methodologies. The first methodology considered a single driving pattern at a time, whereas, the second methodology considered different driving patterns simultaneously for the optimisation. The study considered a simulation model of a power-split HEV for the optimisation of battery size along with internal combustion engine, motor, and generator. An electric-assist charge sustaining supervisory control strategy was considered as the energy management. The maximum speed, acceleration, and gradeability were considered as design constraints. The optimisation was carried out using a genetic algorithm. Fuel economy was considered as an objective for the optimisation. Five standard driving patterns of different traffic conditions and driving styles were considered for the optimisation. Battery system efficiency of each optimum design was calculated over five standard driving patterns. This study found that battery system efficiency of the design which was optimised over different driving patterns was on average 2% higher compared to that of the designs which were optimised over a single driving pattern. This study shows a direction for the selection of battery size in HEVs for real-world application.