Control of the Multi-Input Multi-Output process is a complicated problem due to its non-stationary behavior, coupling, and severe disturbances. The technique of self-tuning the fuzzy PI controller and an analytical technique to determine the scaling factors of Self Tuning Fuzzy Proportional Integral Controller is proposed in this paper. The Multi-Input Multi-Output process considered here is the transfer function of the Heater Mixer Experimental setup. The process is simulated under normal as well as under variations in gain and time constant. The performances of Self Tuning Fuzzy Proportional Integral Controller are compared with the proposed Analytically tuned Self Tuning Fuzzy Proportional Integral Controller for the heater mixer experimental set up in terms of the performance measures Integral Absolute Error (IAE), Integral Square Error (ISE) and Integral Time Absolute Error (ITAE). Simulation results show that the Analytically tuned Self Tuning Fuzzy Proportional Integral Controller gives improved performance compared to Self-Tuning Fuzzy Proportional Integral Controller.