Suicide rates among adolescents have increased dramatically, particularly for Black youth. The majority of suicide decedents have their last clinical contact in primary care. Thus, PPC settings are critical for identifying and treating suicidal youth, but there are challenges with respect to identification, intervention, and implementation. Annual screening for depression using self-report may miss identifying many high-risk youth, as many suicide attempters, particular Black youth, do not report ideation prior to their suicidal behavior and suicidal crises in youth can develop quickly. A second challenge is that once high-risk youth are identified, PPC providers lack a reliable service delivery strategy to effectively treat these youth. A third challenge is that are many barriers for identifying or intervening with Black youth at risk for suicide. Our Signature R01 addresses these challenges as follows: In the first component of the R01, we will develop a predictive analytic platform for PPC based on the electronic health record (EHR), mobile sensing, ecological momentary activity (EMA) assessments of mood and suicidal thoughts and behaviors and self-reports to identify who is at risk and when they are at imminent risk for suicide-related events. To accomplish this, we will recruit 2000 youth from PPC, enriched for those at high suicidal risk, and the sample will be 35% Black. These youth will be followed with interviews and self-reports at 1, 3, and 6 months following baseline and will have 6 months of data from mobile sensing and daily and weekly EMA. We will: (1) develop a predictive algorithm using EHR of adolescents in PPC settings; (2) identify dynamic changes in mobile sensing and EMA measures predicting imminent risk for suicide- related events; (3) develop a data-fusion algorithm combining mobile sensing, EMA, self-reports, and EHR to improve prediction; and (4) test and optimize its performance among Black youth. In the 2nd component, we will conduct a randomized clinical trial (RCT) on a subset of this cohort, namely 900 youth at high suicidal risk. We will compare treatment as usual (TAU) to a suite of tools developed in the current project period to guide the pediatric provider in assessing suicidal risk, making a treatment recommendation, generating a safety plan that is loaded on the patient’s smartphone, and launching an automated texting intervention to increase treatment engagement. Based on our previous work, we hypothesize that this combined intervention, integrated Care to Help At-Risk Teens (iCHART) will decrease suicidal events (suicidal behavior or ideation that results in an emergency referral) by 50%, and the effects will be mediated by increases in referrals, treatment engagement, and safety planning. We will use implementation science methods to assess barriers, facilitators, feasibility, and acceptability of PART predictive analytics and the iCHART intervention to inform future implementation efforts and to promote h...