Abstract. Soil structure is sensitive to intensive soil management. It can be ameliorated by a reduction in soil cultivation and stimulation of plant and microbial mediators for aggregate formation, with the latter being a prerequisite and measure for soil quality. Cover crops (CCs) are part of an integrated approach to stabilize or improve soil quality. Thereby, the incorporation of diverse CC mixtures is hypothesized to increase the positive effects of CC applications. This study entailed an investigation of the legacy effect of CCs on soil aggregates after three crop rotations in the second main crop (winter wheat) after the last CC treatment. Four CCs (mustard, phacelia, clover, and oat) cultivated in pure stands and with a fallow treatment were compared to a mixture of the four CC species (Mix4) and a highly diverse 12-plant-species mixture (Mix12) in a long-term field experiment in Germany. The organic carbon (OC) distribution within macroaggregate fractions (16–8, 8–4, 4–2, 2–1, and <1 mm) and their aggregate stability were measured by dry- and wet-sieving methods, and the mean weight diameter (MWD) was calculated from water-stable aggregates. The results showed that, compared to the fallow, all CCs increased the MWD between 10 % and 19 % in soil under the following main crop. The average MWD increase over the fallow was slightly higher for CC mixtures (16 %) than for single CCs (12 %). Most of the OC (67.9 % on average) was stored in the <1 mm aggregate fraction, highest in the topsoil and decreasing with soil depth. The intermediate fractions (8–4 mm, 4–2 mm, 2–1 mm) stored 8.5 %, 10.5 %, and 11.0 % of the total OC, while 2.1 % was stored in the 16–8 mm fraction. Higher MWD improvement at the 20–30 cm depth also indicates additional benefits from a reduction in the cultivation depth. Structural equation modelling (SEM) suggests that single CCs were more likely to increase OC storage in small macroaggregates <1 mm, while CC mixtures were more likely to increase OC in the largest fraction (8–16 mm). Different individual CC species or mixtures exhibited varying involvement in the formation of different aggregate fractions. We provide evidence that litter quality, root morphology, and rhizosphere input, which affect microbial mediators of aggregate formation, might be the main reasons for the observed differences between CC treatments. Cover crops are valuable multifunctional tools for sustainable soil management. Here, we showed that they contribute to structure amelioration in arable soils. Increasing the functional diversity of plant species in CC mixtures could be a strategy to further enhance the positive effects of CCs in agroecosystems.