Music education in elementary schools continues to face challenges in developing students’ understanding of musical elements, particularly pitch, rhythm, and melody, while deep learning approaches and artificial intelligence (AI)-based technologies offer promising opportunities to improve learning quality, although their implementation in Indonesian primary music education has not yet been systematically reviewed. This study aims to synthesize empirical evidence on the implementation of deep learning-based music education in supporting elementary students’ understanding of musical concepts. A Systematic Literature Review (SLR) was conducted following the PRISMA 2020 guidelines. Literature was collected from Google Scholar, ERIC, Scopus, and Publish or Perish databases covering publications from 1997 to 2026. Of the 100 studies initially identified, 32 articles met the inclusion criteria and were analyzed through thematic synthesis. The findings revealed five major themes: digital technology and multimedia-based learning (31.25%), deep learning and meaningful learning (25.00%), musical elements instruction (18.75%), music education in elementary schools (15.63%), and collaborative learning and student engagement (9.37%). The reviewed studies consistently indicated that deep learning approaches integrated with video-based media, collaborative activities, and AI-supported technologies enhanced students’ understanding of pitch, rhythm, and melody. Learning outcomes showed an average achievement score of 92.5 and a learning mastery rate of 100%. The study concludes that integrating deep learning into music education strengthens students’ mastery of musical concepts while promoting creativity, collaboration, and critical thinking; keywords: deep learning, music education, elementary school, systematic literature review.