Boiler is the main component in steam power plant. Classical control laws such as PID are still widely used to control steam pressure, temperature and drum water level of a boiler in steam power plant because of their good structural understanding, simplicity and better tracking control performance. The combination of challenges such as modeling uncertainties, nonlinear dynamics and parameter variation in the characterized model and external environmental disturbances create difficulties to control the boiler parameters in steam power plant. Although the classical PID controller has good static performance especially for linear time invariant systems, its dynamic performance is weak. It has also poor performance on nonlinear and parametric uncertain systems. By introducing fuzzy as an automatic tuner for the PID gains, it is possible exploit the advantage of both control methods. Fuzzy-PID controller will counteract the parameter variation in the boiler by automatically changing the PID gains. In addition, the fuzzy algorithm in a fuzzy-PID controller handles nonlinearity in the control system which makes it more convenient for nonlinear systems such as boiler system. In this thesis, a MIMO nonlinear boiler model was linearized then decoupled into three SISO models. Then after that, fuzzy-PID controllers were designed for the three SISO models. The three SISO models have three output parameters to be controlled namely, pressure, power and water level. In MATLAB simulations, the fuzzy-PID controller has shown better performance in controlling the boiler parameters than the conventional PID controller. For the three independently controlled parameters which are pressure, flow rate and water level, the responses of the PID controller showed a settling time of 13.5sec, 1.45sec and 12.89sec respectively. In case of fuzzy-PID controller the settling time of the system response greatly decreased to 281msec, 432msec and 354msec respectively. In addition to this, the overshoots of the responses were generally diminished in case of using the fuzzy-PID controller compare to the conventional PID controller.