The surfaces of machined workpieces contain the signature of the manufacturing process. Indeed, the surface inaccuracies can be related to process parameters and machine vibrations. This paper aims at using process monitoring to identify the origin of surface topography defects. Thus, vibration and other process parameter measurements are correlated with profile measurements to establish the relationship between vibrations and surface topography for various process conditions. As each machine and process anomaly is reflected differently in the time and frequency domain, different analyses are applied both to temporal signals such as vibrations and to geometric signals obtained from the micro-geometry of the workpiece. Experimental results validate the diagnostic method and verify that the procedure is effective for the identification of topography defects originated not only from chatter vibrations but also from forced vibration coming from the kinematical chain (unbalance, beating...). The diagnostic method is successfully applied in various machining operations such as turning and grinding.