Coral reef ecosystems are vital for marine biodiversity, coastal protection, and local livelihoods, yet environmental and anthropogenic stressors increasingly threaten them. Accurate and timely mapping of these habitats is essential for effective management and conservation planning. This study aims to evaluate and compare the spatial distribution and extent of coral reef habitats around Liang and Rakit Islands, as well as to assess the classification accuracy of Sentinel-2 Level-2A and the SPOT-7 satellite imagery for the years 2016 and 2021. Both datasets were pre-processed through atmospheric and sun-glint corrections, with SPOT-7 imagery further enhanced by water column correction using the Lyzenga algorithm. Supervised classification in ENVI was used to map four shallow-water habitat classes (coral reef, seagrass, sand, and rubble), and classification accuracy was assessed using confusion matrices to compute Producer Accuracy (PA), User Accuracy (UA), and Overall Accuracy (OA). Across both islands, coral reef and seagrass each occupied roughly one quarter of the mapped shallow-water area, while sand and rubble together accounted for about half of the area, indicating a heterogeneous mosaic of fringing reefs and seagrass meadows interspersed with sandy and rubbly substrates. Maps derived from SPOT-7 imagery produced higher overall accuracies (84% for Liang and 82% for Rakit) than those derived from Sentinel-2A imagery (76% and 70%, respectively). SPOT-7-based classifications consistently delineated larger and more detailed extents of coral reef and seagrass habitats, demonstrating the value of higher-resolution imagery for representing fine-scale features in optically complex coastal environments, whereas Sentinel-2A imagery was more suitable for broad-scale habitat discrimination but tended to underestimate narrow or fragmented patches.