The stable operation of oil-paper insulation power equipment is an important link to ensure the reliability of the power system. However, existing methods have difficulty in diagnosing on the oil-paper insulation ageing. Besides, most of them are suitable for laboratory analysis with a lack of fast and effective diagnose method in actual application. Experimental study about accelerating aging has been done, which is based on IEEE recommended law, to acquire the samples about aging oil- paper paper in different degree. Three types of central Raman spectra, including early, middle and late, were constructed. By using Euclidean distance, chi-square distance and jensen-sinon divergence, respectively, three spectral similarity discrimination methods were used to established a point-wise comparison based Raman spectral diagnosis method for oil-paper insulation aging. Diagnostic analysis of 10 test samples obtained in the experiment, and the diagnostic maximum accuracy was up to 100% (10/10).