Severe aortic stenosis (AS) and mitral regurgitation (MR) are frequently undertreated and characterized by persistent sex, racial and ethnic, socioeconomic, and geographic disparities despite effective valve therapies. Whether automated electronic clinician notification (ECN) alerts improve the evaluation and treatment of AS and MR across health systems is unknown. The purpose of this study was to evaluate whether ECN alerts improve guideline-directed evaluation and treatment of significant AS and MR across multiple health systems. ALERT is a multisystem, cluster-randomized clinical trial including clinicians ordering echocardiograms across 5 U.S. health systems encompassing 35 hospitals between August 2024 and September 2025. Clinicians were randomized 1:1 to receive an ECN alert identifying significant AS or MR with accompanying care recommendations or to no alert with usual care. The primary endpoint was a hierarchical composite of time to surgical or transcatheter valve intervention, followed by time to multidisciplinary heart team clinic evaluation within 90 days, analyzed using the stratified win-ratio method. Secondary outcomes included individual components of the composite. A total of 765 clinicians ordering 2,016 echocardiograms were included. In the win-ratio analysis of the primary endpoint, ECN alert was superior to usual care (win ratio: 1.27; 95% CI: 1.05-1.54; P = 0.007), including higher rates of valve intervention (13.4% vs 9.6%; P = 0.005) and multidisciplinary heart team evaluation (22.7% vs 17.9%; P = 0.005) and shorter times to both endpoint components. Effect sizes were similar in AS (win ratio: 1.29) and MR patients (win ratio: 1.23). No evidence of heterogeneity was noted by valve pathology (Pint = 0.821) or across prespecified subgroups (age, sex, race, social deprivation index, inpatient vs outpatient setting, provider specialty, and rurality; Pint > 0.100 for all) and sensitivity analyses yielded consistent results across modified intention-to-treat, intention-to-treat, and per-protocol populations. In this multisystem cluster randomized trial, automated ECN alerts improved timely guideline-directed evaluation and valve intervention for clinically significant AS and MR. These findings suggest that electronic health record-integrated clinical decision support may represent a scalable strategy to reduce undertreatment and improve access to specialized valve care. (Addressing Under-treatment and Health Equity in AS and MR Using an Integrated EHR Platform; NCT06099665).