To investigate the potential effects of changing precipitation on forest ecosystems, the Throughfall Displacement Experiment (TDE) was established on Walker Branch Watershed, Tennessee, in 1993. Three different throughfall amounts were tested: À33% (DRY); ambient (no change, AMB); and 133% (WET). Throughfall manipulations had no statistically significant effects on total C, N, exchangeable Ca 2 1 , Mg 2 1 , bicarbonateextractable P, or extractable SO 4 2À in soils after 12 years of sustained treatments. Increased K 1 inputs in the WET treatment resulted in relative increases in exchangeable K 1 compared with the AMB and DRY treatments. Soil C, N, and extractable P declined in all treatments over the 12-year study, and the declines in N were inexplicably large. Field observations contrasted with earlier simulations from the Nutrient Cycling Model (NuCM), which predicted greater decreases in exchangeable K 1 , Ca 2 1 , Mg 2 1 , and extractable P in the order WET4AMB4DRY, and no change in C, N, and extractable SO 4 2À . The failure of the NuCM model to accurately predict observed changes is attributed to the lack of mechanisms for deep rooting and the transfer of throughfall K 1 from one plot to another in the model. Measurements of element availability using resin membranes during the final years showed higher values in wet and lower values in dry treatments compared with ambient conditions for mineral N, K, Mn, Zn, and Al, but the opposite for B, Ca, and Mg. In the cases of Ca and Mg, the patterns in resin values were similar to those at the soil exchange sites (greatest in the dry treatment) and appeared to reflect pretreatment differences. This study showed that while longer term changes in soil nutrients are likely to occur with changes in precipitation, potential changes over this 12-year interval were buffered by ecosystem processes such as deep rooting.