Axis Tracking Errors (ATEs) of the active and inactive axis of numerically controlled machine tools are presented as new means of detection of tool wear that forgo expensive sensors or modifications of the machining structure, however, very little has been published about their capabilities or limitations as signal source for monitoring. In this paper the ATEs and cutting forces in drilling tests in two different machine tools, with drills of varying wear levels are measured. The sensitivity to wear is compared by introducing Percent Deviation from New Tool (PDFNT) factor, which is applied to the peak-to-peak values of the signals. While the ATEs are very small in magnitude, they are highly sensitive to wear levels, with PDFNTs reaching to 1000% for some axis. In addition, the standard deviation of PDFNTs calculated in drilling of seven holes with the same tool represents the repeatability of ATEs. The PDFNTs for ATEs are rather repeatable, but less repeatable than the PDFNTs of the axial drilling force. Furthermore it is shown that ATEs of different machine tools have different levels of sensitivity to wear levels which necessitates calibrating of monitoring systems using ATEs for each machine tool separately.