Several methods have been proposed to detect tool wear in milling operation using AE (Acoustic Emission) signals or cutting force signals. However, these methods require additional sensors such as an AE sensor or a dynamometer, which incurs additional costs. For this reason, a simple tool life estimation method based on machining time is used. In this study, a sensor-less tool wear estimation method is proposed. In this method, the parameters required for the cutting force prediction are identified continuously from the spindle motor torque signal, which can be monitored within the computer numerical controlled (CNC) machine. The tool wear progress can be estimated by the continuous change in the identified parameters during milling operation. To identify the parameters continuously, a real-time virtual milling simulation is performed in parallel with a physical milling operation. In the experimental results, it was confirmed that the identified parameter corresponding to the edge force component has linear relationship with the flank wear width of cutting edge. Thus the flank wear can be estimated without any additional sensor.