A detection technology of tool wear based on the maximum entropy of information theory and cross entropy theory was presented.Based on the analysis and treatment of vibration acceleration signal in cutting process,the maximum entropy probability values of vibration acceleration signals in different periods were reckoned.Then the cross entropy of the probability distribution in adjacent periods and the cross entropy of each periods relative to the starting period were calculated.On this basis,the tool wear was judged according to the change of cross entropy.The result shows that with the increase of tool wear,the cross entropy of the probability distribution of the cutting vibration gradually increases in the cutting process.So the change trend of cross entropy of maximum entropy probability distribution of cutting vibration acceleration signal can be used to detect tool wear.