Using BP neural network, a predictive model has been set up between the parameters of spunlaced nonwoven fabric and its two important performances which were the moisture permeability and absorption. The results showed the model was very ideal. Based on this, various parameters corresponding to the optimal values were obtained through computer simulation, that while the viscose content was 80% and the surface density was 58g/m2, the maximum permeability was 6627.3 g/m2. d and the horizontal moisture absorption also reached the maximum of 22.41cm. While the viscose content was100% and the surface density was 68g/m2, the vertical moisture absorption reached the maximum of 19.1429cm. In the end, the PCA was used to integrate the three properties into one target and the optimal parameters were also obtained, which viscose content was 80 % and surface density was 58g/m2 .Besides, according to the method of the computer simulation, a variety of different spunlaced nonwovens also can be obtained to meet different needs, which could be a reference in the future production and experiments.