Adolescent internalizing problems (e.g., depression, anxiety) are common, but fewer than half of affected adolescents access treatment. This need-to-access gap may have been exacerbated during the coronavirus pandemic. To help address this challenge, we evaluated the potential mental health benefits of a digital single-session intervention using an evidence-based approach (problem-solving) and delivered in school-based health classes. In total, 357 US students (M[subscript age] = 12.01; 59.10% female; 11.00% Asian, 12.20% Black, 29.30% Latinx, 0.60% Indigenous, 44.20% White) were randomized to a 30-min, online, self-guided program intended to empower adolescents to solve everyday problems (Project SOLVE), or a matched control program targeting study skills (Project Success). Pre-registered primary outcomes were internalizing symptoms from baseline to three months, and hopelessness and goal-planning confidence from pre- to post-intervention for the full sample. Pre-registered secondary outcomes were internalizing symptoms from baseline to three months among youths with elevated symptoms at baseline (N = 69) and program acceptability. In the full sample, there were significant reductions in internalizing symptoms from baseline to three months for both groups, with greater reductions for SOLVE than Success (p = 0.040; d = 0.17). The elevated-symptom subsample also showed significant reductions in internalizing symptoms for both groups, with greater reductions for SOLVE (p = 0.048; d = 0.40). Both groups reported significantly improved hopelessness and goal-planning confidence, plus high program acceptability. Project SOLVE, a 30-min digital mental health intervention that is easily delivered during class, appears to ameliorate adolescent internalizing symptoms. The brevity, ease of implementation in class, and high ratings for acceptability, enjoyability, and helpfulness suggest potential for dissemination in schools.