Machine tool crack detection is to detect any damages in the tool during machining and to predict the breakage of the tool by identifying the appearance of small cracks in the tool during machining. Machine tool state monitoring is critical for controlling the work piece quality and production continuity in the case of mass production. For assessing tool breakage, machine tool vibration monitoring is a suitable means. The results obtained in FE analysis are validated with experimental data. This work gives a methodology for online crack detection in machine tools. This work presents an experimental technique for measuring modal parameters of a rectangular aluminium plate with cantilever boundary condition using only the output data, with the intention to apply the technique to machine tools. Operational modal analysis is be used for damage detection by determining the depth as well as the position of tool cracks. The results obtained are validated with finite element analysis. To locate the crack, 3D graphs of the normalized frequency in terms of the crack depth and location are plotted. The intersection of these contours gives crack location and crack depth. Out of several case studies conducted the results of one of the case study is presented to demonstrate the applicability and efficiency of the method suggested.