Recently, the U.S. Surgeon General issued an advisory highlighting the lack of knowledge about the safety of ubiquitous social media use on adolescent mental health. For many youths, social media use can become excessive and can contribute to frequent exposure to adverse peer interactions (e.g., cyberbullying, and hate speech). Nonetheless, social media use is complex, and although there are clear challenges, it also can create critical new avenues for connection, particularly among marginalized youth. In the current project, we leverage a large nationally diverse sample of adolescents from the Adolescent Brain Cognitive Development (ABCD) Study assessed between 2019-2020 (N = 10,147, M age = 12.0, 48% assigned female at birth, 20% Black, 20% Hispanic) to test the associations between specific facets of adolescent social media use (e.g., type of apps used, time spent, addictive patterns of use) and overall mental health. Specifically, a data-driven exposome-wide association was applied to generate digital exposomic risk scores that aggregate the cumulative burden of digital risk exposure. This included general usage, cyberbullying, having secret accounts, problematic/addictive use behavior, and other factors. In validation models, digital exposomic risk explained substantial variance in general child-reported psychopathology, and a history of suicide attempt, over and above sociodemographics, non-social screentime, and non-digital adversity (e.g., abuse, poverty). Furthermore, differences in digital exposomic scores also shed insight into mental health disparities, among youth of color and sexual and gender minority youth. Our work using a data-driven approach supports the notion that digital exposures, in particular social media use, contribute to the mental health burden of US adolescents.