The use of medium-resolution photogrammetric-derived Digital Elevation Models (DEMs) to model coastal inundation risk is commonplace in the geosciences. However, these datasets are often characterised by relatively large and loosely defined elevation errors, which can seriously limit their reliability. Post-processed and static RTK dual-frequency GPS data and very high-resolution Terrestrial Laser Scanning DSM data are used here to quantify the magnitude and spatial distribution of elevation error on a 10 km coastal section of a medium-resolution photogrammetric DEM. The validation data are captured at two scales and spatial-resolutions to minimise the risk of spatial bias in the validation results. The strengths and shortcomings of each validation dataset are assessed, and the complimentary value of GPS and Terrestrial Laser Scanning for external validation is demonstrated. Elevation errors highlighted in the photogrammetric DEM are found to be significantly larger than suggested by the data suppliers, with a tendency for the larger errors to occur with increasing proximity to the coastline. The results confirm the unsuitability of the DEM tested for the local spatial modelling of coastal inundation risk, highlighting difficulties that may be prone to occur when similar DEM datasets are used in coastal studies elsewhere.