In the age of remote sensing, particularly with new generation Uncrewed Aerial Vehicles (UAVs), there is a broad spectrum of applications, especially in remote and rapidly changing areas such as the Arctic. Due to challenging conditions in this region, there is a scarcity of detailed spatial studies with data that may be used to estimate changes in glacier volume and geomorphological changes caused by permafrost freeze–thaw cycles. Drone-based Digital Elevation Models (DEM) offer a finer spatial resolution with higher accuracy than airborne and satellite-based products that can be used for acquiring, interpreting, and precisely representing spatial data in broad studies. In this study, we evaluate a UAV-based DEM of two High Arctic catchments, Fuglebekken and Ariebekken, located on Spitsbergen Island. The surveys were carried out in July 2022 using a DJI Matrice 300 RTK drone equipped with a photogrammetric Zenmuse P1 camera. A total of 371 images were taken, covering an area of 7.81 km2. The DEM was created by the Structure-from-Motion technique and achieved a centimetre-level accuracy by overlapping very high-resolution images. The final resolution of the DEM was found to be 0.06 m in Fuglebekken and 0.07 m in Ariebekken, with a horizontal and vertical RMSE of 0.09 m and 0.20 m, respectively. The DJI Matrice 300 RTK drone-based DEM is compared and correlated with the aerial mission of the Svalbard Integrated Arctic Earth Observing System (SIOS) conducted in July 2020 and the satellite-based ArcticDEM acquired in July 2018. This allowed the detection of elevation changes and identification of landscape evolution, such as moraine breaches and coastal erosion. We also highlight the usage of DEM in providing detailed morphometric characteristics and hydrological parameters, such as the delineation of catchments and stream channels. The final products are available at the IG PAS Data Portal.