Soil use intensity controls organic carbon dynamics through different practices, such as fertilizer use, tillage and crop rotations. We evaluated selected treatments from two long-term experiments in Uruguay (Typic Argiudols). Experiment I was established in 1963 at INIA La Estanzuela with conventional tillage. Selected treatments were: Continuous cropping with no fertilizer additions (CS-1); Continuous cropping with N-P fertilization (CS-2); and a six-year crop–pasture rotation with 3 years as CS-2 and 3 years with a grass-legume pasture (CS-5). Experiment II was established in 1995 at INIA Treinta y Tres under no tillage. Treatments were: Continuous cropping with annual forage crops for forage (CC); a short rotation (SR) of 2 years as CC and 2 years with a biannual grass-legume pasture; a long rotation (LR) of 2 years as CC and 4 years with a grass-legume pasture; and a permanent pasture (PP) overseeded every four years. Soil organic carbon (SOC) and organic C in particulate organic matter (C-POM) were determined. After 45 years SOC stocks (0-20 cm depth) in experiment I declined 45 and 29% in CS-1 and CS-2 compared to CS-5. Inclusion of pastures (CS-5) resulted in significantly higher values of C-POM than continuous cropping systems and the relative magnitude of these differences were higher than those observed with SOC. A higher C input through residue return from crops and pastures, as well as less soil erosion may explain these findings. In Experiment II, after 11 years, SOC (0-15 cm depth) declined 23% in CC compared to PP due to a higher biomass removal. Moreover, C-POM was more sensitive to detect differences between CC and PP. Crop-pasture rotations (SR and LR) reduced SOC (11 %) compared to PP. Regardless of the type of tillage, the inclusion of perennial pastures is necessary to maintain or increase SOC in the long term.