The study proposes a new UCR program that differs from the existing program in two fundamental ways. First, rather than sending only monthly summary statistics to the national program, State or local agencies will submit individual records for each incident and arrest that occurred during the month. Second, two reporting levels will be established; most contributors will provide basic offense and arrest information similar to that currently reported, but a smaller sample of agencies will report much more extensive information. All large agencies will be expected to participate in the second reporting level. This two-level reporting meets the needs for increased depth and scope of regional and national statistics about crime. The proposed new system also includes improvements in procedures for ensuring the quality of UCR data, an expanded series of publications, enhanced analysis capabilities and user services, and better compatibility with National Crime Survey data and Offender-Based Transaction Statistics. Appendixes contain a description of study methodology, a discussion of the relationships between the National Crime Survey and the UCR Program, an estimated cost of the proposed UCR system by task, and a sample design for the level II component. A bibliography with approximately 225 listings is provided.