The invention relates to a paper money identification technology, and particularly to a method for true and false identification of paper money through main component analysis. As artificial anti-fake point identification is adopted in the prior art, image shift is difficult to avoid, the obtained true and false identification features are limited, the stability and the true and false identification capability cannot reach to the requirement, and the identification accuracy rate is low. The method for true and false identification of paper money through main component analysis comprises the steps of: taking true paper money and false paper money samples for sample training so as to obtain sample features and projection matrixes; enabling the testing money samples to enter into a multi-spectral money-counting machine, a counter and an automatic teller machine to compare features and matching the similarity with the sample features and the projection matrixes so as to judge whether the paper money is true or false. The method for true and false identification of paper money through main component analysis has the advantages that manual work is unnecessary for finding the features in the identification process, the extracted features have high distinguishing degree for the true and false paper money, the transportability is very good as a machine can have a rapid true and false identification capability after a training result is transplanted to a program, the true and false identification capability is strong, and the method is applicable to true and false identification of paper money of various editions of various countries in the world.