Principal component analysis is the well-known method in pattern recognition, but classical principal component analysis extract some features that keep maximal scatter and the algorithm doesn't use the classificatory information of samples. Therefore, extracted features aren't very efficient to classification based on classical principal component analysis. Based on the image retrieve principle, the paper presents a kind of retrieve space principal component analysis (RS-PCA). Then, a supervised retrieve space principal component analysis (SRS-PCA) using classificatory information are developed according to RS-PCA. The algorithm makes the extracted features more effective and the recognition precision is increased. The experiments resulted on ORL and Yale face database demonstrate that the proposed algorithm has more powerful and excellent performance than classical principal component analysis.