Background: Type 2 Diabetes, a common, complex condition with high comorbidity, affects 24% of Veterans. Vulnerable Veterans (e.g., African-American, rural, comorbid mental health diagnosis, low-income) are less likely to have controlled diabetes, and have higher mortality and morbidity compared to other Veterans. Significance/Impact: Health Care Informatics interventions to support chronic disease self-management through technology can improve access, health equity, and health outcomes for vulnerable Veterans. Customizable, interactive self-management support through Annie addresses the VA's priority of improving access, including via virtual modalities, and providing a tailored experience that incorporates Veteran needs and preferences (VA Strategic Imperative 2). Research priority areas of access to care, women's health, mental health, primary care practice, informatics, virtual care, health equity, and patient-centered care are all addressed. Innovation: By incorporating the needs and preferences of vulnerable Veterans with diabetes in a self- management texting intervention and testing its effectiveness against a more traditional education-only intervention, lessons learned can improve the development of text-based support for other complex chronic conditions. It can also improve future implementation of Annie-based self-management support throughout VA. Specific Aims: AIM 1: Refine and beta test components of an interactive, tailored self-management texting protocol (DD-TXT) using a participatory design process incorporating vulnerable Veterans' preferences, VA clinician input, and evidence on effective texting-enabled self-management programs. The DD-TXT protocol will consist of: Core Messaging: Customizable core modules on medication management, blood sugar and blood pressure monitoring, preventive care, problem solving, appointment reminders, administrative messages; and Optional Messaging: A library of patient-selected modules (e.g. nutrition, physical activity, weight management, emotional coping, goal setting) designed to motivate and educate. AIM 2: Conduct a randomized controlled comparative effectiveness trial with 400 Veterans whose diabetes was uncontrolled (defined as HbA1c over 8.0% for at least 50% of the most recent 6 months) in 2018 in Gainesville, FL or Chicago, IL. The primary aim will be to assess the comparative effectiveness of DD-TXT compared to DSE, a diabetes skills education-only texting protocol based on a skills workbook that is currently given to VA patients with diabetes. The primary outcome will be HbA1c percent time in control. Secondary outcomes include self-reported adherence to diabetes self-care recommendations (SCI-R), diabetes self- efficacy, diabetes distress, LDL, and blood pressure control. We hypothesize that DD-TXT will result in better proximal health outcomes and diabetes self-management behaviors vs an education-only protocol (DSE). AIM 3: Obtain information to guide future implementation of diab...