Integration of topology data of the entire road ahead into the powertrain control unit in combination with hybrid electric powertrains offers a huge potential for optimising the control strategy. The intension of the present paper is to show a methodology for computing the maximum potential of fuel saving over a specified route. For this offline calculation a non-linear state-space model of the longitudinal dynamics is used to find the fuel optimal trajectory using Bellman’s Dynamic Programming (DP). As the model is of third order with three control inputs and DP can only be applied to systems with a maximum sum of inputs and states of four, the iterative variant of DP is used. This special method described in this paper allows the solution of optimisation problems with a number of states and inputs up to six, without leading to insufferable computing times.