Prediction of the behavior of T-shaped chambers due to their high complexity has always been of great interest to researchers. In this article, based on experimental data and genetic programming, the optimal model was presented for mixing process response. To obtain the system’s behavioral equations, first, using the experimental results and by changing the input variables system, input – output data is extracted. In order to predict the behavior of the system, the equation of input – output data, is derived using genetic programming. To design the structure of genetic programming trees, multi-objective optimization with two objective functions is taken into consideration: model inaccuracy and complexity of structure. By minimizing the objective function at the same time, we are looking for simple equations (minimizing the complexity of the structure) and increasing the accuracy of modeling (minimizing the error). In order to achieve a less complex equation, depth of the generated trees in structure of genetic programming will be minimal. By using multi-objective optimization, optimum set of points have been presented. Comparing the results obtained from the models and real data represents a very good match.