An approach helpful when developing an optimized construction of a 6/4 type switched reluctance motor (SRM) is described in the paper. The analytical modeling procedure, based on the reluctance network method and analytical solution of an ordinary differential equation, enables applying a gradient optimization routine and better control of optimization process. The model allows for estimation of the efficiency, torque, and acoustic noise of the motor taking into account the magnetic non-linearity and the control algorithm to keep a constant input power. A bicriterial optimization routine has been applied to find optimal constructions. Eleven geometric and winding parameters are supposed to be the optimization quantities. Analyzed constructions – the initial one and the optimized ones, were validated by means of FEM calculations. The proposed approach can be employed in designing the SRM to be a drive motor in an electrical vehicle, at least as a first attempt.