This paper presents a methodology for optimisation of vehicle drivetrain configuration and their design with respect to multiple parameters. Some preliminary results from such optimisations are presented. Multi-objective optimisation algorithms have been developed to improve on earlier work on "environomic" optimisation, in which costs and environmental parameters are optimised using an agglomerated objective function. The vehicle component models are based on a combination of experimental data and theory, and in some cases neural net approximations of complicated subsystems have been used to replace slow theoretical models. The components developed at LENI have been incorporated into the vehicle simulation system ADVISOR and the system has been optimised with respect to component sizes, as well as three configurations (one conventional and two hybrid drivetrains) for such diverse objectives as emissions, costs and performance.