Pattern recognition techniques are used to learn to automatically recognize physical objects (e.g., alphanumeric characters) or abstract multidimensional patterns (e.g., n points in a d-dimensional space). The recognition could involve (i) classification into one of pre-specified categories (supervised learning), (ii) grouping or clustering (unsupervised learning), or (iii) learning the similarity between two patterns, say face images. Pattern recognition systems are routinely used for license plate recognition, document classification, speech and speaker recognition, blood cell(leukocyte) classification, military target recognition, industrial inspection, etc. The design of a pattern recognition system requires the following modules: (i) sensing, (ii) feature extraction, (iii) learning decision rules, and (iv) performance evaluation. In this talk I will introduce the topic of forensic pattern recognition and provide examples of systems for fingerprint matching, face recognition and ballistic image matching based on breech face impression.