Accurate and detailed land cover information is essential for effective biodiversity monitoring and conservation efforts. In the Caribbean Netherlands (consisting of Saba, St. Eustatius, and Bonaire), this need is particularly pressing, as the islands face growing ecological pressures from climate change, invasive species, and human activity. However, the islands’ limited financial resources, steep terrain, and dense vegetation make systematic field surveys difficult, restricting the collection of consistent and detailed information about land cover. Remote sensing offers a powerful alternative, but previous satellite-based efforts for Saba were limited by their spatial and radiometric resolution, as well as the lack of vertical structural data, resulting in maps that lacked ecological specificity and taxonomic resolution. In early 2024, high-resolution airborne Light Detection and Ranging (LiDAR) and passive multispectral optical imagery were collected over the islands for the first time, providing a unique opportunity to assess their potential for land cover mapping. The objective of this study is therefore to investigate the potential for mapping land cover of the Caribbean Netherlands using the recently acquired airborne LiDAR and passive optical data. Specifically, this study investigates two main aspects, focused on Saba: (1) the quality of the LiDAR-derived Digital Terrain Model (DTM) released with the dataset, and (2) the combination of LiDAR structural parameters with optical remote sensing data for land cover mapping. To assess the quality of the LiDAR-derived DTM, a novel data-driven reliability algorithm was developed. This algorithm combines several properties of the LiDAR data indicative of DTM reliability into a single pixel-based DTM reliability score RDT M , which indicates how trustworthy the DTM is. Applying this method to Saba shows that 42% of the island has zero DTM reliability (no ground coverage at all), 9% low reliability, 16% moderate reliability, and 33% high reliability. Comparison of these reliability classes with vegetation indices confirmed that areas with dense vegetation tend to have lower DTM reliability. In addition, several LiDAR-derived structural parameters were identified that describe both the topography and vertical structure on the surface, such as vegetation. These parameters were combined with products from passive multispectral optical satellite imagery to evaluate their potential for land cover mapping in two case studies. The first case study examined how LiDAR structural information can reveal variation within a single land cover class from a previous study, demonstrating that distinct vegetation structures can be distinguished within what was previously mapped as uniform forest. The second case study investigated whether structural parameters can explain the occurrence of the invasive vine Coralita (Antigonon leptopus), showing that its topographic occurrence was consistent with the literature. The results demonstrate that the newly acquired airborne LiDAR dataset provides valuable structural information for land cover mapping. However, challenges remain in areas with dense vegetation and steep terrain, where limited ground penetration reduces DTM reliability. Furthermore, with the integration of field validation data (which was unavailable), the dataset’s full potential for land cover mapping could be more fully realized.