A common strategy for mixture composition analysis based on Raman spectroscopy is to construct the spectral database of pure substances and calculate the spectral similarity of the mixture to be identified and the pure substances in the database. However, influenced of the performance of the measuring instrument and the mutual interference of the components of the mixture, the spectrum of the substance contained in the mixture to be identified will have different degrees of distortion compared with the corresponding pure substance spectrum in the database, bringing great difficulties in component identification. To address this problem, a method to improve the identification accuracy of components in mixtures using the spectral data of known mixtures is proposed herein. The spectral peak information, including Raman shift and full width at half maximum, of the pure substance in the database is used to study the correspondence of the spectral peaks of the known mixture to the specific substances of the mixture. The spectral feature parameters of the pure substance, the known mixture, and the mixture to be identified are constructed using the Raman shift of the spectral peak, the full width at half maximum, and the peak intensity, respectively and the fuzzy membership function is used to calculate the spectral similarity between the mixture to be identified, the pure substance, and the substances contained in known mixture based on calculated feature parameters. Furthermore, suspected components contained in mixture to be identified are determined based on the spectral similarity. Based on the spectral database of 204 pure substances and 8 known mixtures, the experimental results for 81 unknown mixtures reveal that the proposed method can reduce the calculation error of similarity caused by spectral distortion and can improve the identification accuracy. Compared with search strategy based on the database of pure substance, the identification accuracy obtained using the proposed method increases from 76.34% to 92.83%.