A fingerprint feature extraction algorithm based on wavelet transform was proposed. Firstly, the paper centered on the core-points, then divided the fingerprint image into an effective area. Next, the area was analyzed by two-dimension wavelet transform, and the energy of every passage was accurately extracted as the fingerprint features. The proposed algorithm required less computational effort than conventional algorithms which were based upon minutia features extraction. In addition, this algorithm did not need the high quality fingerprint image. Besides, the correct recognition rate also reached a high level.