This study reviews how Digital Twin (DT) a virtual replica dynamically linked to physical systems via real-time data has been applied in education. A systematic literature review (SLR) was conducted on Scopus, IEEE Xplore, and Google Scholar, covering 2020–2025. Searches were performed on [month year] using a predefined query, followed by deduplication and a two-stage screening (title/abstract and full text). From 79 identified records, 20 studies met the inclusion criteria. The synthesis shows three dominant domains: (1) DT-based immersive and safe training for engineering/TVET (virtual machines and learning factory), (2) smart campus facility management using IoT-driven DT for energy, health, and space optimization, and (3) Digital Twin of the Learner for adaptive learning and early-risk prediction. Across studies, DT is consistently reported to reduce training risk and operational cost while improving authenticity of practice-based learning, although evidence strength varies by context and evaluation design. Key adoption barriers include infrastructure cost, interoperability, and cyber-physical security. DT therefore holds promise for a connected educational ecosystem, provided institutions start with targeted pilots and strengthen data governance and teacher capacity.