This thesis is devoted to apply a Model Predictive Fuzzy Controller (MPC and Takagi-Sugeno) to a specific Steam Boiler Plant. This is a very common problem in control. The considered plant is based on the descriptions obtained from the data of a referenced boiler in the combined cycle plant as Abbot in Champaign, Illinois. The idea is to take all the useful data from the boiler according to its performance and capability in different operation points in order to model the most accurate plant for control. The considered case study is based in a modification of a model proposed by Pellegrinetti and Bentsman in 1996, considering to be tested under the demands of the Control Engineering Association (CEA). The system is Multi-Input and Multi-Output (MIMO), where each controlled output has a specific weight in order to measure the performance. The objective is to minimize cost index but also make it operative and robust for a wide range of variables, discovering the limits of the plant and its behaviour. The model is supposed to manage real data and was constructed under real physical descriptions. However, this model is not a white box, so the analysis and development of the model to be used with the MPC strategy have to be identified to continue with the evaluation of the controlled plant. There are some physical variables that have to be taken into account (Drum Pressure, Excess of Oxygen, Water Level, Water Flow, Fuel Flow, Air Flow and Steam Demand) to know if these variables and other parameters are evolving in the correct way and satisfy the logic of the mass and energy balances in the system. After measuring and analysing the data, the model is validated testing it for different values of steam demands. The controller is tuned for every one of the considered demands. Once tuned, the controller computes the manipulated variables receiving information from the controlled ones, including their references. Finally, the resulting controller is a combination of a set of local controllers using the Takagi-Sugeno approach using the steam demand setpoint as scheduling variable. To apply this approach, a set of local models approximating the non-linear boiler behaviour around a set of steam demand set-points are obtained and then their a fused using the Takagi-Sugeno approach to approximate any unknown steam demand located in the valid range of values.