Biodiversity experiments generally report a positive effect of plant biodiversity on aboveground biomass (overyielding), which typically increases with time. Various studies also found overyielding for belowground plant biomass, but this has never been measured over time. Also, potential underlying mechanisms have remained unclear. Differentiation in rooting patterns among plant species and plant functional groups has been proposed as a main driver of the observed biodiversity effect on belowground biomass, leading to more efficient belowground resource use with increasing diversity, but so far there is little evidence to support this. We analyzed standing root biomass and its distribution over the soil profile, along a 1-16 species richness gradient over eight years in the Jena Experiment in Germany, and compared belowground to aboveground overyielding. In our long-term dataset, total root biomass increased with increasing species richness but this effect was only apparent after four years. The increasingly positive relationship between species richness and root biomass, explaining 12% of overall variation and up to 28% in the last year of our study, was mainly due to decreasing root biomass at low diversity over time. Functional group composition strongly affected total standing root biomass, explaining 44% of variation, with grasses and legumes having strong overall positive and negative effects, respectively. Functional group richness or interactions between functional group presences did not strongly contribute to overyielding. We found no support for the hypothesis that vertical root differentiation increases with species richness, with functional group richness or composition. Other explanations, such as stronger negative plant-soil feedbacks in low-diverse plant communities on standing root biomass and vertical distribution should be considered.