Tree-based land-use systems such as agroforestry are expected to have better carbon (C) sequestration potential than most agricultural systems. Silvopastoral systems that integrate trees into animal production systems are the most widely practiced form of agroforestry in North America. C sequestration potential of such silvopastoral systems, especially the stable C pool in the soil, have, however, not been studied. We hypothesize that silvopastoral systems will have more stable (recalcitrant) C fraction in soil than under tree-less pasture under similar ecological settings. Soil C pools were examined in four farms of slash pine (Pinus elliottii) + bahia grass (Paspalum notatum) silvopastural systems in two soil types (Spodosol and Ultisol) in northern Florida. The specific objectives are to compare the distribution of labile and recalcitrant forms of C among different soil fractions (sand, silt, and clay), at different depths in silvopasture, treeless-pasture, and local natural vegetation systems. Soil samples were drawn from positions between trees in a row (TS), on an alley in a silvopasture (AS), on treeless-pasture (TP) and on natural vegetation (NV) systems at six different soil depths (0 – 5, 5 – 15, 15 – 30, 30 – 50, 50 – 75, and 75 – 125 cm). Plots at each site had similar management history, and by assumption, similar soil C levels prior to introducing silvopasture. Soils will be fractionated into sand, silt, and clay and the labile and recalcitrant forms of soil C in each fraction determined. The results will be used to evaluate the influence of slash-pine-based silvopastoral system on stable soil C, and thus C sequestration potential, in the two soil types. The work is in progress; results will be presented at the meeting.