In this paper, three model predictive control (MPC) strategies are developed and implemented for coal-fired power plant cycling applications. These strategies correspond to a dynamic matrix control (DMC)-based linear MPC, the proposed nonlinear MPC (NLMPC) and a nonlinear MPC from the literature (L-NMPC) as benchmark. For application purposes, dynamic models of two different coal-fired power plants with carbon capture are addressed: an integrated gasification combined cycle power plant with a water-gas shift membrane reactor (IGCC-MR) and a supercritical pulverized coal-fired power plant with monoethanolamine-based post-combustion carbon capture (SCPC-MEA). Successful MPC implementations on IGCC-MR system and SCPC-MEA carbon capture subsystem are addressed, including cycling trajectory tracking and disturbance rejection scenarios. The closed-loop results show that the proposed nonlinear MPC (NLMPC) improves the control performance by up to 96% when compared to the DMC controller in terms of the integral squared error (ISE) results. (C) 2018 Elsevier Ltd. All rights reserved.