Fire risk assessment in forest stands relies on detailed information about the availability and spatial distribution of fuels. In particular surface fuels such as litter, downed wood, herbs, shrubs and young trees determine fire behaviour in temperate forests and constitute the primary source of smoke emissions. Remote sensing has been suggested as a potentially valuable tool to estimate the spatial distribution of fuels across large areas. However, accurately estimating surface fuel loadings in space across various fuel components using airborne or spaceborne sensors is complicated by obstruction from the forest canopy. In addition, mapping efforts have largely focused on simplified representations of fuel situations for specific modelling purposes, such as classifications into fuel types or fuel models, rather than estimating fuel loadings. In this work, we test whether the fusion of high-resolution LiDAR data (> 60 points/m²) with moderate- to high-resolution satellite imagery from the Sentinel-2 mission (10-20 m) allows to predict loadings of all surface fuel components using machine learning techniques. Our analysis is based on a field inventory of surface fuels in a mixed temperate forest with two dominating deciduous tree species (Fagus sylvatica, Quercus petraea) and two dominating coniferous species (Pinus sylvestris, Pseudotsuga menziesii). We produce fine-scale maps of surface fuel loadings that can form the basis for fuel management strategies as well as for calculations of fire behaviour characteristics and fire effects. Furthermore, we test how spatial variability in surface fuel loadings is captured when broader categories such as fuel types are used as mapping units. We investigate possible relationships of overstory tree species and cover with surface fuel loadings to reach more general conclusions about predictors for surface fuel loadings in temperate forests of central Europe. Our study contributes to a better understanding of fuel-related fire risk in temperate forests, which can help in developing appropriate forest management decisions and fire-fighting strategies.