To maintain public safety and improve system response, a strong need exists for understanding the characteristics of individuals involved in gun crimes, related police activities, and situational/contextual variables surrounding gun incidents. A serious knowledge gap exists, however, in understanding weapon-related offenses at the city level and in how to develop a local guncase database. This paper, which is based on research experiences at an urban police department in USA, describes and analyzes the process and related issues in developing such a database. The study revealed a few critical lessons concerning this process, including those related to communication and cooperation between the police and researchers, quality of data, and importance of a refined dataset for improving local police policies and operations.