Most methods for detecting fingerprint reference point need to scan the image or orientation field pixel by pixel or block by block to determine whether it is a candidate reference point, which is inevitably time-consuming.A novel fast algorithm is proposed in this paper to improve the efficiency by walking directly to the singular region without scanning the fingerprint image.Then in the local area around the singular region, an enhanced algorithm for detecting fingerprint reference point based on mean-shift concept is proposed and utilized to locate the precise position of the reference point.Experimental results on FVC2000 DB1a and DB2a databases validate that the proposed method outperforms the others extraordinarily in terms of efficiency and accuracy.