Milk cooling process is one of the most important milk processing steps after milk extraction process. Milk cooling process is carried out to inhibit the microbial growth. The milk is cooled from room temperature to 4 °C. Generally, this process is performed in batch installation. In this research, continuous milk cooling would be simulated. First, the model of the system was derived from the mass and heat balance. The simulation results were then identified using the System Identification Toolbox (SIT) on Matlab. System Identification Toolbox was used to build the mathematical models of dynamic system based on measured input and output data. The result of system identification is useful for setting control and stability analysis. The milk cooling system was then controlled by Proportional Integral (PI) Control. There are two kinds of tuning methods that will be analyzed. These are Hagglund-Astorm method and Tyreus-Luyben. The results showed that the control performance tuned using the Tyreus-Luyben method was better than the Hagglund-Astorm method from the criteria of its SSE (sum squared of error).