Abstract: A mathematical model with waterborne polyurethane (WPU) content, temperature during the curing process and the reduction degree of microfiber leather as variables was developed to predict the thermal and humidity comfort performance of microfiber suede leather by performing orthogonal experimentation, using both statistical regression and neural network methods. The model was subjected to normality testing and analysis of variance, and the influencing factors were analyzed and compared to assess the predictive ability of the model. The results showed that the WPU content had a significant effect on moisture permeability, and the degree of reduction had the greatest effect on breathability. The fitting coefficients of breathability and moisture permeability in the statistical regression prediction model were 0.871 and 0.865, respectively, while these coefficients in the neural network prediction model were 0.997 and 0.989, respectively. Overall, the neural network model performed better than the statistical regression model in predicting the thermal and wet comfort of microfiber suede leather.