Linear Model Predictive Control (MPC) continues to be the technology of choice for constrained multivariable control applications in the process industry. Successful deployment of MPC requires “getting right” multiple aspects of the problem. This includes the design of the regulatory controls that receive setpoints from MPC, design of the multivariable controller(s) themselves, test design for model identification, model development, and dealing with nonlinearities. In the following, we highlight approaches and techniques that are successfully applied in practice and provide an overview of recent technological enhancements that are being made to MPC. While significant progress has been made in both the technology and practice, there are challenges with MPC, mostly related to the effort required to develop an application and to ensure adequate performance over time. Suggestions for addressing these issues are included as possible research directions.