Ethanol currently remains the leading biofuel in the transportation sector, with special focus on spark ignition engines, as a pure as well as a blend component. In order to provide reliable numerical simulations of gasoline combustion processes under the influence of ethanol for modern engine research, it is mandatory to develop well validated detailed kinetic combustion models. One key parameter for the numerical simulation is the laminar burning velocity. Under the aspect of minimizing the general simulation effort for burning velocities, well-validated models have to be reduced. As a base kinetic mechanism for the reduction and optimisation process with respect to burning velocity calculations, a detailed model presented by Zhao et al. (Int. J. Chem. Kin. 40 (1) (2007) 1-18) is chosen. The model has been extensively validated against shock tube, rapid compression machine and burning velocity data. The detailed model consists of 55 species and 290 reactions. A stochastic model calibration approach is undertaken for the optimisation of the base mechanism against data found in the literature. New experimental data at 5 bar and 373 K are used for validation. The optimised mechanism is significantly reduced within this work applying a multi-stage reduction strategy using the directed relation graph with error propagation (DRGEP) technique. The reduced mechanism is again validated with the experimental data used before. Overall, the reduced mechanism consists of 36 species and 215 reactions. It predicts experimental flame speeds very well.