Abstract South African young women (YW) have annual HIV incidence rates of 5-10% in recent HIV prevention trials. Pre-exposure prophylaxis (PrEP) is a strategy with potential to provide high protection against HIV if adherence is high. Two key questions about PrEP delivery to African young women need to be assessed: 1) how to identify and motivate at-risk YW to use PrEP, and 2) how to efficiently support their PrEP adherence. For PrEP uptake, we will evaluate a provider-driven approach based on behavioral and clinical risk assessment through the VOICE risk score, and a patient-directed approach to encourage patients to assess their risks and motivations for PrEP using a decision support tool. We hypothesize that a patient-centered decision tool will result in higher uptake of and adherence to PrEP than a provider-driven approach alone. For PrEP adherence, it is unknown what level and type of adherence support is needed for African YW; some women may be highly adherent with standard brief counseling or two-way SMS, in which case resources would be saved by only using more intensive adherence support for women who are not adherent. We will use a sequential multiple assignment randomized trial (SMART) design to evaluate a stepped PrEP adherence support model that begins with brief counseling, or two-way SMS, both of which are readily scalable, and then assesses the effectiveness of two strategies for more intensive adherence support (drug level feedback and cognitive behavioral counseling) only in those not adherent after 3 months of PrEP use. We hypothesize that if YW are motivated to use PrEP, they will adhere well with simple support through brief counseling, reminders with triage. Those who need additional adherence support will respond to drug level feedback and monthly counseling to address adherence challenges. Our Specific Aims are: Aim 1. Using a randomized design, test whether use of a patient-facing decision- support tool in addition to a provider-administered risk score positively influences PrEP initiation compared to use of a risk score alone. Aim 1a: Evaluate if use of the decision tool affects YW’s risk perception and PrEP initiation; Aim 1b: Determine if the decision tool alters YW’s decision-making and providers’ counseling about PrEP. Aim 2. Test, using a SMART design, a stepped model of scalable adherence support strategies in South African young women who initiate PrEP. We will randomize 330 women ages 16-25 in Johannesburg who initiate PrEP to standard of care brief counseling with or without two- way SMS. Adherence will be assessed at month 3 using tenofovir levels to categorize women as ‘responders’ and ‘non-responders.’ Non-responders will undergo secondary randomization to quarterly counseling about their drug levels or monthly CBT counseling. The primary outcome will be on PrEP with drug detected at 12 months. In this project, we will determine optimal strategies for facilitating PrEP uptake and scalable adherence support ...