The chapter provides an overview of several recent applications of model predictive control (MPC) to coordinated power and thermal management in connected and automated electrified vehicles. Such applications of MPC face significant challenges due to the need to accommodate different timescales of faster power and slower thermal dynamics, interactive constraints, and the need to generate and accommodate long-term vehicle speed and traction power forecasts in the optimization. It will be shown that a multi-horizon formulation of MPC consisting of a shorter receding horizon phase and a longer shrinking horizon phase can help address these challenges. Perspectives on creating forecasts and exploiting multi-horizon MPC for energy efficiency optimization based on several case studies are summarized.