The dynamic characteristics of the NC machine are the important technical indicators of NC machine performance, and the precision of existing machine's dynamic characteristics evaluation analysis that based on structural dynamics modeling and simulation is poor. For this problem, this paper presents NC machine dynamic characteristics evaluation methods based on Grey Correlation Analysis. This method, by measuring the vibration signal of feature points in NC machine and making 1/3 octave spectrum analysis for vibration signal, can obtain the characteristics data sequence of vibration signal. Through the grey correlation analysis method, calculate the relevancy between the characteristics data sequence of each vibration signal and that of ideal dynamic characteristics, and quantitatively evaluate the dynamic characteristics of machine tools according to the size of the grey correlation. Took three same type NC milling machines as study objects, designed the cutting vibration resistance experiments, used this paper's method for the evaluation analysis of the comprehensive dynamic characteristics of NC milling machine, and obtained the evaluation results which is in line with the actual dynamic characteristics of machine tools, and verified the evaluation method of comprehensive dynamic characteristics that proposed in this paper.