Fingerprint identification is broadly established as one of the most reliable identification methods for forensic and non-forensic applications. Nevertheless, due to growth in volume of fingerprint databases, performance requirements, and speed limitations of the manual process, Automated Fingerprint Identification Systems (AFIS) are extensively needed. This paper describes the design and implementation of a cost-effective AFIS system that operates in four stages: enhancement, classification, feature extraction and matching/identification. This system has been tested on two sets of multiple fingerprint images acquired using an inked impression method. These tests have confirmed system accuracy and reliability for forensic applications.