Published open access in Remote Sensing, the article evaluates hyperspectral mapping of seagrass and macroalgae habitats at Remøy on the Norwegian west coast, using two SeaBee platforms: an aerial drone with a hyperspectral imager (10 × 10 cm resolution) and a surface drone carrying an underwater hyperspectral imager (1 × 1 cm resolution). Spectral angle mapper classification reached 89% and 86% overall accuracy for the two platforms, and the results show how they complement each other: the aerial drone delivers broad spatial coverage, while the centimetre-scale surface data provides high-quality training and validation data, reducing the need for field-based ground truthing.