Release or uptake of soil carbon has the potential to affect atmospheric carbon dioxide concentrations and hence feedback to greenhouse gas forced climate change. We conducted extensive observations of soil carbon cycling in three montane meadows spaced at elevation intervals (~300 m) that effect average temperature variations in the range expected under a doubled CO2 climate (~2 C). We find that carbon in the top 10 cm of soil can be explained (R2~0.7) by a simple function of plant productivity, litter quality, and soil microclimate that is derived from a steady-state model of carbon pools and flows. Because the variables used in this function are easily observed, the validity of the function can be tested for other unsaturated ecosystems, and if successful, used as part of efforts to predict soil carbon mediated feed-backs to climate change. Future research to explore soil carbon response to land use in the context of Californias historic forestry and agricultural practices will be outlined.