The accurate personal recognition is important and critical concept in a wide range of applications such as computer login, physical access control, notional ID card, and criminal investigations. Biometric represents the most secure method to recognize personal identity and from all the biometric technologies fingerprint is the most reliable and accurate one. The objective of this thesis is to develop and implement a fingerprintbased biometric system which is capable of recognizing individuals with high level of confidence and minimum error rate. An efficient image enhancement algorithm is used to improve the clarity of the fingerprint ridges and valleys and facilate the extraction of minutiae points through the minutiae extraction stage which is performed using the crossing number method. The reliable and fast extraction of minutiae proves the accuracy and the reliability of the crossing number method. To find the mated minutiae between two fingerprint images, the alignment based matching algorithm is implemented with a number of modifications to increase system reliability and decrease matching error. The algorithm has the ability to find the correspondence between the minutiae points efficiently and adaptively compensates the nonlinear deformations and inexact transformations between an input and a template. The designed system is implemented using MATLAB V 7.0 and tested on a set of fingerprint images gathered from 20 different persons. The experimental results show that the proposed system has high verification rate (98.9%) and low equal error rate (2.6%). The obtained results prove the reliability and accuracy of the designed system which is able to recognize the individuals efficiently and with very low error rate.