Three-dimensional track reconstruction is a key issue for directional Dark Matter detection. It requires a precise knowledge of the electron drift velocity. Magboltz simulations are known to give a good evaluation of this parameter. However, large TPC operated underground on long time scale may be characterized by an effective electron drift velocity that may differ from the value evaluated by simulation. In situ measurement of this key parameter is hence a way to avoid bias in the 3D track reconstruction. We present a dedicated method for the measurement of the electron drift velocity with the MIMAC detector. It is tested on two gas mixtures : CF4 and CF4 + CHF3. We also show that adding CHF3 allows us to lower the electron drift velocity while keeping almost the same Fluorine content of the gas mixture.