Understanding the effect of planting densities and species proportions on light absorption and light use efficiency can help to improve the management of mixed-species forest plantations. Our study aimed to disentangle the role of light interception and light use efficiency (LUE) on the biomass production of Eucalyptus grandis (E), a highly productive species in tropical conditions, and Acacia mangium (A), a N2-¿xing species, in monocultures and mixed-species plantations for contrasting planting densities. A randomized block experiment set up over 4 ha in southern Brazil was intensively monitored for 14 months at mid rotation. The absorbed photosynthetically active radiation (APAR) was simulated for each tree of the experiment using the tri-dimensional MAESPA model parameterized with detailed in situ measurements of tree and foliage. LUE for stem wood production was estimated as the ratio of measured stem biomass production (SBP) and simulated APAR. The APAR of Eucalyptus trees did not significantly differ between monocultures and mixed plantations, the reduction of Eucalyptus density being compensated by an increase in light absorption of Eucalyptus individuals. The LUE of Eucalyptus trees in monoculture and mixed-species stands was found to be comparable only at low planting densities. The replacement of Eucalyptus trees with Acacia trees resulted in a reduction in Eucalyptus LUE only at high planting density. The SBP of Eucalyptus trees was mainly explained by differences in APAR, while both APAR and LUE explained the SBP of Acacia trees. The maximum stand production was obtained with monoculture of Eucalyptus at high density and no mixture reached this productivity. Reducing the proportion of Eucalyptus in mixture lead to a substantial decrease in stand production at high planting density due to a decrease in LUE, while this stand production reduction was offset at low planting density, underlying a higher diversity effect at low planting density. In the perspec