As a result of soil management systems, like the conventional tillage system (CTS) versus the no-tillage system (NTS) for vegetable crops, modifications in soil properties occur, which change, for example, the rates of soil aggregation and the content of total organic carbon (TOC). The aim of this study was to quantify the TOC contents and evaluate the rates of soil aggregation and the distribution of aggregates by diameter classes in a soil planted to onion in NTS and CTS and compare them to an adjacent forest area in Ituporanga, Santa Catarina, Brazil. We evaluated the following treatments, which consisted of planting cover crops, alone and intercropped, in NTS: spontaneous vegetation (SV); 100 % oats; 100 % rye; 100 % wild radish; intercropping of wild radish (14 %) and rye (86 %); and intercropping of wild radish (14 %) and oats (86 %). We also evaluated an onion-growing area under CTS for +/- 37 years, and a forest area (secondary forest for +/- 30 years), both adjacent to the experimental area. In September 2013, five years after setting up treatments with cover crops, undisturbed soil samples were collected from the 0-5, 5-10, and 10-20 cm soil layers, and aggregates were separated to assess stability by the wet method. In the aggregates, we quantified TOC and, after separation into diameter classes (8.00 mm>empty set >= 0.105 mm), we calculated the mean weighted diameter (MWD) and mean geometric diameter (MGD) of the aggregates, the distribution of the aggregates in macroaggregates (empty set >= 2.0 mm), mesoaggregates (2.0>empty set >= 0.25 mm), and microaggregates (empty set<0.25 mm), and the aggregate sensitivity index (SI). Data underwent analysis of variance and principal component analysis (PCA). The highest TOC contents were found in the forest area (52.83, 37.77, and 26.70 g kg(1) for the depths of 0-5, 5-10, and 10-20 cm, respectively) and the lowest in the CTS (18.23 g kg(-1), 0-5 cm). Treatments with cover crops, alone or intercropped, did not differ among themselves (p <= 0.05) for TOC, or in relation to the control area (SV). The CTS had the lowest rates of MWD (3.425, 3.573, and 3.401 mm), MGD (2.438, 2.682, and 2.541 mm) and SI (0.77, 0.79, and 0.81) in the three layers evaluated. For MWD and MGD, differences were not observed (p <= 0.05) among treatments in NTS, but these rates were higher than the rates of CTS; the MWD readings were equal to the forest area; and the MGD readings were higher in the 0-5 and 5-10 cm layers. In the 10-20 cm layer, in the NTS, treatment with wild radish showed higher values of MWD (4.520 mm), MGD (4.284 mm), and SI (1.07). Regarding the distribution of aggregates by diameter classes, the CTS showed the smallest (14.22, 14.75, and 13.86 g) values for macroaggregates, and largest (4.94, 3.44, and 3.2 g/3.0, 3.0, and 3.76 g) values for meso-/microaggregates, respectively. The NTS showed higher macroaggregate values (means of 19.90, 20.48, and 18.56 g) compared to the forest area (16.0, 16.31, and 15.47 g) and to the CTS (14.22, 14.75, and 13.86 g) in the three layers evaluated. The use of cover crops, alone or intercropped, in NTS of onion was quick in recovering and increasing the TOC and the MWD, MGD, and SI indices in relation to the CTS, and, compared to the forest area, the MGD increased (0-5 and 5-10 cm). Oilseed radish increased soil aggregation (MGD and SI) in the 10-20 cm layer compared to the other treatments with cover crops.