Tool wear in machining changes the geometry of the cutting edges, which affects the direction and amplitudes of the cutting force components and the dynamics in the machining process. These changes in the forces and dynamics are picked up by the internal encoders and thus can be used for monitoring of changes in process conditions. This paper presents an approach for the monitoring of amulti-toothmilling process. The method is based on the direct measurement of the output from the position encoders available in the machine tool and the application of advanced signal analysis methods. The paper investigates repeatability of the developed method and discusses how to implement this in a process monitoring and control system. The results of this work show that various signal features which are correlated with tool wear can be extracted from the first few oscillating components, representing the low-frequency components, of the machine axes velocity signatures. The responses from the position encoders exhibit good repeatability, especially short term repeatability while the long-term repeatability is more unreliable.