In an andisol tt was placed for three years land uses systems like: farmer traditional (slash and burn, fertilization and sowing of bean and maiz); Quesungual Agroforestry System-SAQ (no burn, natural regeneration of native trees, bean and maiz sowing); crop residues as management cover, and a fi ve years secundary forest (tacotal) as a control. It was considered soil samples at 0-5, 5-10 y 10-20 for physical and fertility characterization and field measurement of soil erosion by rain simulation. Quesungal and crop residues systems, remained soil erosion susceptibility at levels as low as the natural secondary forest.