A method that was originally developed for the investigation of microstructures has been adapted to the measurement of rough technical surfaces. Based on a combination of angular resolved light scattering measurement with a data analysis by linear multivariate regression, almost any surface parameter can be estimated. This includes height parameters such as the Ra and Rz roughness as well as spatial parameters such as the autocorrelation length T. To this end, a linear model is generated from several well-studied samples and some kind of a priori information. This model can then be used to calculate the surface parameters from the angular distribution of the scattered light. A generalized solution for the inverse scattering problem cannot be obtained, but the method is specially suited to measure deviations from rated values and to detect deviations in the machining process. In this paper, the basic principle of the method, a concept for the description of the angular distribution of scattered light, and the utilized data analysis technique are presented. Furthermore, the results of experiments that were made with honed surfaces may demonstrate its applicability in process control.