In an attempt to understand better the local biogeochemistry of the South China Sea (SCS) and to unravel the contribution of this marginal low-latitude basin to changes in atmospheric CO2 concentrations, we analyzed the carbon isotopic composition of organic matter (&dgr;13Corg) in four sediment cores from throughout the SCS covering the last 220 kyr. Higher values (around -19.5 to -20.5?) mark glacial stages, while lower values (around -21 to -22.5?) are characteristic of interglacials. Following well established procedures, the &dgr;13Corg records are converted to local pCO2 estimates. On the basis of these and other low-latitude &dgr;13Corg-pCO2 estimates from the literature, we present a critical evaluation of the use of &dgr;13C of bulk sedimentary organic matter to hindcast past changes in local CO2(aq). Three crucial pitfalls are identified. (1) Given the present inability to quantify precisely the time-varying amount of terrigenous Corg input to marine sediments, absolute values of pCO2 estimates based on bulk sedimentary Corg are questionable. (2) None of the low-latitude sedimentary &dgr;13Corg-pCO2 records shows the expected correlation between temporal changes in upwelling intensity and CO2 estimates, most likely due to the antagonistic influences of CO2(aq) and phytoplankton growth rate on &dgr;13Corg. (3) A detailed comparison of marine &dgr;13Corg-pCO2 records with the Vostok CO2 record reveals significant differences in phasing, specifically at the end of the last deglaciation and during the oxygen isotope stage 5/4 transition. However, in areas where equilibrium between oceanic and atmospheric CO2 occurs, for example the SCS and the Mediterranean, the timing of changes in &dgr;13Corg should agree with the CO2 record from ice cores if &dgr;13Corg is a reliable proxy for changes in CO2(aq). Taken together, the compilation of records presented here cautions the use of &dgr;13Corg as an unambiguous tracer of dissolved molecular CO2 in the surface ocean and calls for a re-evaluation of the role of the low-latitude ocean on temporal changes in atmospheric CO2. ¿ 2001 American Geophysical Union