Despite the efficacy of psychotherapies, ~90% of people with alcohol use disorder (AUD) do not receive treatment for this chronic condition. Novel virtual care approaches (telephone, video, portal therapies) conceptually rooted in efficacious therapies [e.g., motivational interviewing (MI), cognitive behavioral therapy (CBT)] and theory can potentially improve treatment accessibility and appeal and reduce stigma. Virtual approaches have had limited use in AUD care, but recently increased to provide ongoing care (telephone/video therapy) for current patients in specialty clinics, while treatment initiation remains low. There is an untapped potential for virtual approaches to engage individuals across health systems, who otherwise do not receive treatment, but could benefit. Virtual strategies, including telephone and video sessions and patient portals, are promising for reaching and engaging substance-using populations within health systems. Using a portal messaging system and phone calls to virtually engage patients in AUD treatment outside of clinics and in their preferred locations is an innovative treatment model that can potentially be implemented in health systems, but must be tested to inform broader dissemination. We will use a sequential, multiple assignment randomized trial (SMART) to identify critical adaptive intervention (AI) strategies for a virtually-delivered AUD engagement and care model. Adults with AUD will be randomized to a 1st stage strategy for drinking reduction and AUD care engagement integrating referral for AUD care: 1) a single telephone MI session (T- engage), or 2) 4-weeks of a MI-focused portal messaging (P-engage). At 4 weeks, non-responders will be randomized to a 2nd stage strategy: 1) step up to a video 8-session MI-CBT for AUD (phone delivery as needed), or 2) continued 1st stage (i.e., 4 weeks of P-engage with greater depth, a second T-engage session building on the first). Outcomes will be assessed at 4-, 8-, and 12-months. Aim 1