Understanding how forest biodiversity and carbon stocks can be simultaneously maintained or enhanced is crucial for the development of effective forest management strategies and policies aimed at halting biodiversity loss under climate and land use change. Many existing policies and strategies assume that forests hosting high carbon stocks provide co-benefits in terms of high biodiversity. While this relationship has been demonstrated in tropical forests, it has not been rigorously assessed in temperate forests, and further research is needed to understand how carbon stock and biodiversity covary, and whether non-linearities, community thresholds and variability across taxa exist. We built a comprehensive biodiversity dataset including six taxonomical groups (beetles, birds, bryophytes, fungi, lichens, and vascular plants) distributed over 352 plots representing three forest types, in 22 distinct study areas across three European countries (France, Hungary and Italy). We used Boosted Regression Trees and Threshold Indicator Taxa ANalysis (TITAN) to address three questions: (1) is there a positive relationship between overall multi-taxa diversity (multidiversity) and aboveground carbon stocks in European temperate forests? (2) Are there nonlinearities or thresholds (e.g. synchronous change-points in species composition) along the carbon stock gradient we considered, and are they consistent across taxonomical groups and forest types? (3) How do individual species respond to increasing aboveground carbon and how does the proportion of species with positive (win-win species) or negative (trade-off species) relationships vary across taxonomical groups? We found neither higher multidiversity, nor a consistently higher number of species across the six taxonomical groups in forests storing higher aboveground carbon. We found evidence for synchronous change-points in community composition along the carbon gradient, although these were neither consistent, nor equally abrupt across taxonomical groups and forest types. For all taxonomical groups except lichens, trade-off species markedly decreased at carbon stocks between 80 and 110 Mg/ha, roughly corresponding to the canopy closure phase of the forest stand development. The turnover between trade-off and win-win species was gradual along the carbon gradient, and the proportion of these two groups of species varied across taxonomical groups, and forest types. Maximizing forest carbon storage may not benefit all facets of biodiversity at the same time, although it may enhance some specific taxa and functional groups. Explicitly distinguishing between win-win and trade-off species, and considering taxon-specific community change-points along the carbon gradient could provide a conceptual basis for forest managers to operatively reconcile biodiversity conservation and carbon stock retention.