Aim: Estimating the current spatial variation of biomass in the Amazon rain forest is a challenge and remains a source of substantial uncertainty in the assessment of the global carbon cycle. Precise estimates need to consider small-scale variations of forest structures resulting from local disturbances, on the one hand, and require large-scale information on the state of the forest that can be detected by remote sensing, on the other hand. In this study, we introduce a novel method that links a forest gap model and a canopy height map to derive the biomass distribution of the Amazon rain forest. Location: Amazon rain forest. Methods: An individual-based forest model was applied to estimate the variation of aboveground biomass across the Amazon rain forest. The forest model simulated individual trees; hence, it allowed the direct comparison of simulated and observed canopy heights from remote sensing. The comparison enabled the detection of disturbed forest states and the derivation of a simulation-based biomass map at 0.16 ha resolution. Results: Simulated biomass values ranged from 20 to 490 t (dry mass)/ha across 7.8 Mio km(2) of Amazon rain forest. We estimated a total aboveground biomass stock of 76 GtC, with a coefficient of variation of 45%. We found mean differences of only 15% when comparing biomass values of the map with 114 field inventories. The forest model enables the derivation of additional estimates, such as basal area and stem density. Main conclusions: Linking a canopy height map with an individual-based forest model captures the spatial variation of biomass in the Amazon rain forest at high resolution. The study demonstrates how this linkage allows for quantifying the spatial variation in forest structure caused by tree-level to regional-scale disturbances. It thus provides a basis for large-scale analyses on the heterogeneous structure of tropical forests and their carbon cycle.