Diabetes disease management programs (DDMP) are proliferating, but their overall impact in improving quality of care using Health Employer Data and Information Set (HEDIS) quality metrics has not been well studied. Furthermore, DDMPs are usually ongoing, but the incremental benefits of continuing the program beyond the initial patient educational intervention have not been rigorously tested. This study evaluates the impact of length of DDMP participation on diabetes-related HEDIS 2002 quality indicators across 20 health plans. Results are stratified by duration of DDMP participation into three levels, "full participants" (6-12 months duration), "partial participants" (<6 months duration) and "non-participants" (0 months duration). The overall national compliance rate across all six combined HEDIS quality measures was 65.6% among full-participants (FP), 58.4% among partial-participants (PP) and 57.0% among non-participants (NP). This study demonstrates that participants in a comprehensive DDMP fair better than non-participants and that those with sustained participation (>6 months) benefit the most.