This paper presents the design of a predictive speed controller for electrical drives. The control scheme does not have a cascaded structure; instead, it uses a single optimization algorithm to generate the control action to be applied at the next sampling instant. The control is based on the optimization of a cost function, which is designed using the insight given by the solution of the minimum-time control problem for a simplified system, namely, the double integrator, which is a reasonable model of the machine dynamic behavior. The experimental results obtained with a permanent-magnet synchronous motor and an induction machine, fed by a two-level three-phase inverter, confirm that this approach reaches a very good dynamic performance without a high computational effort.