This paper proposes a novel fingerprint matching technique, which matches the fingerprint minutiae by using an effective representation to capture both the local and global minutiae features and their relative directions in fingerprint according to these minutiae orientations. Exploiting this representation, we characterize the local features of each minutia which describe a rotation and translation invariant features in its close sub-region. The matching algorithm used the minutiae covering zones (sub-regions) technique to compute the global matching. They are used to find the correspondence of two minutiae sets and the global fingerprint features at the same time. Thus, our approach doesn?t embed fingerprint alignment into the rotation of the feature vector stage to overcome the critical reference point location and orientation problem as defined in the original approach [4]. The matching algorithm was tested on a subset of the fingerprint database DB2 used in FVC2002 and the results are promising.