Abstract: A multi-objective genetic algorithm was used to optimize the structure of microchannels with circular cavities and internal ribs. The objective functions of thermal resistance and pumping power were established based on CFD simulation results by response surface methodology (RSM), and a multi-objective genetic model was developed to optimize the structure parameters of microchannels. Pareto front of thermal resistance and pumping power was calculated by non-dominated sorting genetic algorithm NSGA-Ⅱ and representative solutions were obtained by k-means clustering. Moreover, the comprehensive heat transfer performance in microchannels was evaluated using thermal enhanced factor. The calculation results show that the multivariate statistical coefficients R² of thermal resistance and pumping power objective function are 0.932 9 and 0.996 6, respectively, which means that the fitting has high accuracy. The more uniform temperature distribution could be obtained with higher thermal enhanced factor of 1.23 in the microchannel with optimized structure. The comprehensive heat transfer performance in the microchannel with higher thermal resistance or pumping power is worse than that in the microchannel with moderate thermal resistance and pumping power.