Surface serves as the fingerprint of a component. Most researchers understood surface topography from a single-stratum viewpoint, some ones have focused on a bi-Gaussian stratified model that respects formation mechanism, but few studies have considered a multi-Gaussian stratified possibility. A multi-Gaussian stratified model of surface topography is developed based on the existing bi-Gaussian stratified surface theory. The bi-Gaussian characterizing method, on the basis of linear regression, is revised to differentiate the mixed multiple Gaussian components. The model and the method are demonstrated on simulated multi-Gaussian stratified surfaces and real engineering surfaces manufactured by turning, rough lapping, and fine lapping. The results reveal the existence of multi-Gaussian stratified feature on multi-process surfaces in industry, and improve the stratified surface theory.