Printing adaptabilities like quantitativeness, smoothness, whiteness, gloss and roughness of different inkjet printing paper were tested in order to study the feasibility of predicting ink jet printing quality by generalized regression neural network. Chroma was measured after printed under the same printing conditions. And a prediction model was established by utilizing generalized regression neural network combined with printing chroma and printing printability of ink jet printing paper. As predicted by the prediction model based on generalized regression neural network, the minimum average color difference of the GRNN model was 4.7215, and the average maximum color difference was 4.8638. The results show that the model can quantitatively describe the effect of ink jet printing paper printability on chromatic aberration of printing products and provide experimental and theoretical basis for the selection of paper.