Model predictive control (MPC) differs from other control methods mainly in its implementation of a control action. Usually, MPC solves a finite-horizon optimal control problem at each sampling instant, so that the control actions for the current and a period of future time are obtained. However, only the current control move is applied to the plant. This work explains the how this useful model permits the flexibility of acting while optimizing. It details a variety of models for one and two-step control, explains state feedback synthesis approaches with and without a finite switching horizon, and open and closed-loop optimization, as well as output feedback synthesis approaches.