To detect soil changes related to vegetation and fertility restoration in a long fallow agricultural system of the Venezuelan Andes, 32 soil (A horizon, 0 -15 cm depth) and litter characteristics were studied in plots at different stages of the fallow-cultivation cycle. Four sectors of the valley were sampled, each one including seven plots: recently ploughed after a long fallow period; 1 and 2 years under potato crop; 1, 4 -5 and 8 years in fallow and natural vegetation. Each sector had similar topography, parent material and exposure to reduce the spatial heterogeneity that can hinder the synchronic analysis of the succession. Data of each sector were standardized before the statistics comparisons. Although all soils were acidic, those involved in the cropping cycle, or with only 1 year of fallow, had a significantly lower pH than the others, indicating that soil cultivation triggered off acidifying processes intense enough to overcome the strong buffering indices of the soils. These acidifying processes, facilitated by the acidity generating ions that widely dominate the desaturated CEC, are surely due to the stimulation of litter and soil organic matter mineralization after ploughing the soil and also to the N fertilization of the cultivated soils. The high soil contents of exchangeable Al 3þ and free Al oxides suggest that Al plays an important role in SOM stabilization, lowering its mineralization. No successional increase of any main plant nutrient was found in soil or litter. Moreover, soil available P and litter-P contents are higher during the cultivation phase and at the beginning of the succession, probably as a consequence of fertilization. Water holding capacity was similar for all soils, indicating that this long fallow agriculture system does not change the water storage capability of the soil, an important aspect for the role of the páramo in the regional water balance. The main characteristics of soils and litters, including their d 13 C values, were useless to monitor soil changes during cultivation and fallowing. Nevertheless, the d15 N values of SOM decreased steadily along the crop-fallow chronosequence, while those of litterfall were rather constant from recently ploughed to 1 year fallow soils, decreasing suddenly on medium fallow plots and again on the virgin páramo. Together, the d15 N values of soil-and litter-N grouped the soils following the crop-fallow chronosequence, suggesting that a change from 'open' to 'closed' N cycling is the characteristic that better discriminates the soils along the succession. q