Lava domes form by the effusive eruption of viscous lava and are inherently unstable and prone to collapse. Dome collapses can generate pyroclastic flows and trigger explosive eruptions and thus represent a significant natural hazard. Many processes may contribute to the instability and collapse of lava domes, including advance of the dome margins, overtopping of confining topography, internal gas overpressure, and gravitational instability of the dome structure. Collapses that result from these processes can generally be grouped into two types: active and passive. Active collapses are driven by processes associated with active lava effusion, (e.g. dome growth or gas pressurization), while passive collapses are not directly associated with eruptive activity and can be triggered by overtopping of topographic obstacles or weakening of the dome structure. We use data collected by uncrewed aerial systems (UAS, commonly called ‘drones’) and a slope stability model to both identify and assess the stability of potential collapse sites for both passive and active processes. We collected visual and thermal infrared images by UAS and used structure-from-motion photogrammetry to generate thermal maps and digital elevation models (DEMs) of two example lava domes at Sinabung Volcano (Sumatra, Indonesia) and Merapi Volcano (Java, Indonesia). We evaluate the stability of erupted lava using the Scoops3D numerical model to assess the risk of passive and active collapses, including an assessment of the effect of lava material properties and internal pore pressure on the dome stability. We compare the collapse risk from Scoops3D with UAS-derived temperature maps and DEM differencing to evaluate the stability, size, and location of observed or potential collapses. We test whether Scoops3D can hindcast the sites and magnitudes of passive collapses at Sinabung that occurred in 2014 and 2015 and assess the stability of the remaining lava dome (growth has ended in spring 2018). For both volcanoes. Through application of these techniques, we are able to evaluate the collapse risk due to multiple processes that may act contemporaneously to generate dome instability. This study demonstrates how identification and classification of individual collapse mechanisms can be used to assess hazards at dome-forming volcanoes.