Terrestrial laser scanning (TLS) has been used extensively in Earth Science for acquisition of digital outcrop data over the past decade. Structure-from-motion (SfM) photogrammetry has recently emerged as an alternative and competing technology. The real-world performance of these technologies for ground-based digital outcrop acquisition is assessed using outcrops from North East England and the United Arab Emirates. Both TLS and SfM are viable methods, although no single technology is universally best suited to all situations. There are a range of practical considerations and operating conditions where each method has clear advantages. In comparison to TLS, SfM benefits from being lighter, more compact, cheaper, more easily replaced and repaired, with lower power requirements. TLS in comparison to SfM provides intrinsically validated data and more robust data acquisition in a wide range of operating conditions. Data post-processing is also swifter. The SfM data sets were found to contain systematic inaccuracies when compared to their TLS counterparts. These inaccuracies are related to the triangulation approach of the SfM, which is distinct from the time-of-flight principle employed by TLS. An elaborate approach is required for SfM to produce comparable results to TLS under most circumstances.