The objective of this study was to evaluate soil chemical attributes in a soil conservation experiment for onion, under different cropping systems. The study was conducted on a Humic Dystrudept in Ituporanga, Santa Catarina state, Brazil. Eight different cropping systems for onion were evaluated: T1: maize-onion succession, T2: common vetch-maize/rye+fodder raddish-onion-maize/rye+fodder raddish-common bean, T3: rye-onion-maize/black oat-maize, T4: onion-velvet bean succession, T5: rye-onion-millet/black oat-onion-millet, T6: rye-onion-velvet bean succession, T7: onion-velvet bean+millet+sunflower succession, from T1 to T7 were conducted under no tillage system (NTS), T8: maize-onion succession, was conducted under conventional tillage system. We used the randomized complete block design, with 5 replications. All the chemical samples were collected and evaluated at the 0-5, 5-10 and 10-20 cm soil layers, the soil chemical attributes evaluated as follows: total organic carbon (TOC), particulate organic carbon (POC), carbon associated to soil minerals (CASM), N and P. At the 0-5 cm layer, P concentrations were severely affected by the different cropping system. In terms of treatments similarities, the treatments may be grouped in three clusters as follows: cluster 1: T2, T3, T7, cluster 2: T1 and T8 and cluster 3: T4, T5, T6. For N concentration, the different cropping system affected its concentration at the 10-20 cm layer. Carbon associated to soil minerals wasn´t affected by the different cropping systems, indicating that more time is needed to study this soil organic carbon fraction. Keywords: crops rotation, nutrients, cover crops, clustering.