In spite of evidence for positive diversity-productivity relationships increasing plant diversity has highly variable effects on the performance of individual plant species, but the mechanisms behind these differential responses are far from being understood. To gain deeper insights into the physiological responses of individual plant species to increasing plant diversity we performed systematic untargeted metabolite profiling on a number of herbs derived from a grassland biodiversity experiment (Jena Experiment). The Jena Experiment comprises plots of varying species number (1, 2, 4, 8, 16 and 60) and number and composition of functional groups (1 to 4; grasses, legumes, tall herbs, small herbs). In this study the metabolomes of two tall-growing herbs (legume: Medicago x varia; non-legume: Knautia arvensis) and three small-growing herbs (legume: Lotus corniculatus; non-legumes: Bellis perennis, Leontodon autumnalis) in plant communities of increasing diversity were analyzed. For metabolite profiling we combined gas chromatography coupled to time-of-flight mass spectrometry (GC-TOF-MS) and UPLC coupled to FT-ICR-MS (LC-FT-MS) analyses from the same sample. This resulted in several thousands of detected m/z-features. ANOVA and multivariate statistical analysis revealed 139 significantly changed metabolites (30 by GC-TOF-MS and 109 by LC-FT-MS). The small-statured plants L. autumnalis, B. perennis and L. corniculatus showed metabolic response signatures to increasing plant diversity and species richness in contrast to tall-statured plants. Key-metabolites indicated C-and N-limitation for the non-leguminous small-statured species B. perennis and L. autumnalis, while the metabolic signature of the small-statured legume L. corniculatus indicated facilitation by other legumes. Thus, metabolomic analysis provided evidence for negative effects of resource competition on the investigated small-statured herbs that might mechanistically explain their decreasing performance with increasing plant diversity. In contrast, taller species often becoming dominant in mixed plant communities did not show modified metabolite profiles in response to altered resource availability with increasing plant diversity. Taken together, our study demonstrates that metabolite profiling is a strong diagnostic tool to assess individual metabolic phenotypes in response to plant diversity and ecophysiological adjustment.