Early, accurate identification of potentially at-risk students is crucial to developing effective dropout prevention programs. Checklists are commonly used by schools and agencies to identify potential dropouts. Unfortunately checklists have several weaknesses, e.g., it is difficult to know how many characteristics must be checked before deciding a student is at risk. Despite their shortcomings, checklists can be helpful as initial screening devices. An identification system is more than just an instrument, it is a process. Suggested processes to help a school district develop its own system include: data collection process; data analysis process; and data utilization process. Although variables related to dropping out include those from several sources, such as the family and community, ones for which schools can collect data, and to which they can respond include: attendance, grade point average, number of grade retentions and discipline referrals, educational level of parents, special program placements, free/reduced lunch program, number of school transfers, reading and math scores, and language spoken in home. With caution and continuous modification and refinement, school districts can use variables from extant checklists, the research literature and local surveys as the basis for developing their own identification systems, especially computer-generated predictive formulas. By using a more systematic approach, school administrators can be more confident in student assignment to programs, and practitioners can be more effective in designing relevant strategies for target students. (LLL)