Reducing greenhouse gas emissions to alleviate global warming will certainly be one of the major challenges of the 21st century. Transportation plays a very important part in this, which is why the European Commission and the European manufacturers have found an agreement to limit the average emissions of vehicles to 130 gCO₂/km in 2012 and 95 gCO₂/km in 2020. Cutting vehicles' consumption of hydrocarbons is becoming a critical issue to reach these ambitious targets. Electric vehicles, characterized by zero direct CO₂ emissions, seem to be a relevant way to achieve these CO₂ emissions.Despite their capabilities to emit no local pollution and to operate silently, electric vehicles have also one important drawback: the limited autonomy offered to the customer. As for conventional vehicles, energy consumption for electric vehicles is very dependant of driving conditions, such as driving cycles and ambient temperature operating conditions for instance. Consequently, assessment of the electric vehicle autonomy can be very dependant of assumptions taken into account during its design, which may be a real limitation to the use and deployment of these types of vehicles.The objective of this paper is to present a global methodology to design an electric vehicle and to evaluate the sensitivity of this sizing process to the assumptions and the requirements taken into account. The presented methodology consists in transforming vehicle and customer requirements towards objectives and constraints on the electric powertrain. Several types of vehicle requirements and driving cycles are taken into account. A tool able to design the electric machine and to evaluate its efficiency has been used in this study. Finally, the design phase leads to a final electric powertrain adapted to the whole requirements while optimizing its global efficiency and its dimensions.