Fluvial wood is recognized as one of the key elements of riverine environments. Its importance, role, and potential related risks are widely considered and analysed. Still, to characterize and measure wood elements in the field, a large investment in terms of time and cost is normally required. These activities are often carried out in complex and potentially dangerous environments. In this context, the use of Unmanned Aerial Vehicles (UAVs) provides a powerful alternative technology to traditional approaches. In this work, we propose a novel approach combining UAV surveys with Structure from Motion (SfM) photogrammetry to quantify fluvial wood. The proposed methodology was tested in the Blanco River (southern Chile), which was severely affected by the Chaitén volcanic eruption in 2008 that delivered large quantities of wood to the channel. An area of about 1.2 ha was surveyed using UAV and field measurements. A comparative analysis between the traditional field-based and UAV surveys was carried out to quantify the accuracy of the latter and to better understand the strengths and potential limitations of the proposed methodology. Results showed that differences between the two methodologies were about 13% for both single large wood elements and wood jam volumes. Moreover, a sensitivity analysis applied to 12 different scenarios allowed us to define the measurement precision achieved using the orthomosaics and DEMs. Our results showed that the proposed approach is effective in terms of field and laboratory activities, saving around half the time. This study demonstrates that the combination of UAV and SfM can yield accurate results, covering large areas in less time and providing a powerful tool for performing repeated surveys and thus acquiring relevant information for wood budgets and wood dynamics in rivers.