ABSTRACT People who use drugs (PWUD) with acute medical problems have high rates of subsequent mortality and morbidity related to substance use, with notable racial disparities. While harm reduction services to reduce overdose and injection-related complications have historically functioned outside of health care settings, integrating harm reduction into acute care clinical settings could improve these outcomes. There is a critical gap in how best to implement harm reduction services in the health system that will lead to effective behavioral change and address equity gaps. Our study proposes to develop an acceptable, feasible, and effective peer-led bundle of harm reduction services. Our overall objectives are to tailor the THRIVE (Teaching Harm Reduction In Vulnerable Environments) intervention and determine its efficacy in changing behaviors and reducing health risks among PWUD. The THRIVE intervention was conceptualized by our team along with people with lived experience and informed by the COM-B model for Behavior Change and the Theoretical Domains Framework. THRIVE bundles evidence-based harm reduction strategies (e.g. safer injection education, take home naloxone, and fentanyl test strips). It includes a face-to-face session boosted by weekly text messages and electronic content. Content is delivered over 12 weeks. Our research team has the extensive expertise in community-partnered research, qualitative methodology, behavioral health intervention design, and clinical trial management needed to successfully complete the proposed aims. We will use a human-centered design approach to tailor the THRIVE model and address the patient, provider, and systems-level barriers to implementation in hospital and emergency-department settings. This includes a “Design Sprint” in which PWUD will map the problem (guided by qualitative interviews with patients and healthcare providers), sketch implementation elements, choose the key aspects to develop, and build model components. These implementation methods for the THRIVE model will then be user-tested with 20 patients over an 8-week period. We will then examine the efficacy of the THRIVE intervention in reducing the cumulative incidence of self-reported non-fatal overdose or skin and soft tissue infection between baseline and 6-months among PWUD in a hybrid type 1 randomized controlled trial. We will conduct a randomized controlled trial recruiting patients (n=390) admitted to the hospital or Emergency Department with opioid use disorder in one of three hospitals. We will also analyze implementation barriers and facilitators of the THRIVE model using the health equity implementation framework to identify any needed implementation supports for widescale implementation.