Cases are confirmed and supplemental information is obtained from the Colorado Uniform Crime Report/Supplemental Homicide Report, police investigations, medical records, and newspaper accounts. Electronic data sets are created that contain the information manually extracted from these sources. An evaluation of Colorado's firearm injury surveillance system compared system- identified cases with injuries and deaths identified from external sources such as hospital trauma registries, emergency department logs, and paramedic trip reports. This evaluation determined the system is flexible, acceptable to reporting agencies, and sensitive in that it detected 69 to 77 percent of firearm-related hospitalizations identified from external sources. In addition, the system's predictive positive value is high, with 89 percent of potential cases confirmed as actual cases. The system is simple in design but requires significant staff time to abstract information. Further, because existing data sources are used, timeliness of the data is often delayed. Additionally, minor injuries (those requiring less than 24 hours or no hospitalization) are not detected. Despite limitations, the system is a valuable resource for information about firearm injuries and deaths and represents an important first step in reducing the number and severity of firearm injuries in Colorado. 10 references, 2 tables, and 1 figure