Data on plant communities (biomass and relative cover of all target species), plant traits (41 different traits, measured on 59 species), and 42 ecosystem properties/functions, measured between 2003 and 2012 in the Jena Main Biodiversity experiment. In floodplain grasslands of the Saale river, near Jena (Germany) 78 20x20 m grassland plots were set up, in which combinations of 1, 2, 4, 8 or 16 species were sown, from a species pool of 60. Thereby, the aim was to create a gradient in plant species richness and functional composition. In each year from 2003-2012, relative cover (in %) of each target species was estimated within 3x3 m subplots. In addition, plant biomass was measured in both spring and summer.In addition, we compiled trait data for 59 of the 60 sown species, based on a combination of existing literature, pot experiments and measurements in the Jena Main Biodiversity experiment monoculture (1-species) plots. Data on 41 traits was collected. Finally, we measured in 41 different ecosystem functions in the Jena Main Biodiversity experiment. Each ecosystem function was measured in at least 3 different years between 2003 and 2012.The "R2.model.random.text[x]" (where x is a number from 1 to 40) are secondary data files, and the outcome of statistical models. In these, 100 times a random subset of 1 to 40 (out of the 41) plant traits were analysed as predictors of the 42 ecosystem functions, in order to assess how the proportion of variance in ecosystem functioning explained by traits (R2 values) depends on the number of traits analysed.