In pelagic carbonate sediments, the degree to which the δ13C values of inorganic and organic fractions co-vary has been used to interpret rates of production, burial and decomposition of organic carbon. This relationship is relatively consistent through time, permitting estimates of organic carbon production and preservation. However, as the majority of pelagic sediments older than 200 Myr have been subducted, carbonate sediments deposited in epeiric seas and platforms are often substituted for pelagic carbonates in analyses of ancient global carbon cycling. There are well-known pitfalls to using shallow marine carbonate materials, including diagenesis, semi-isolation of depositional environments and input of different types of sediments with varying inorganic δ13C (δ13Cinorganic) values, which can obscure any global signatures. One method used to assess whether global changes in δ13C are accurately represented by δ13Cinorganic records is to examine variations in the δ13C of co-occurring organic material (δ13Corganic). If a δ13Corganic record co-varies with a co-occurring δ13Cinorganic record, it is argued that the signals must be related to variations in the global carbon cycle. This assumption has been investigated by analysing the isotopic composition of the organic carbon preserved in the uppermost 150 m of periplatform sediments recovered during ODP Leg 166 from the western margin of Great Bahama Bank. The δ13Corganic values measured in this study were compared to previously published δ13Cinorganic records measured on identical samples, thus allowing a study of the correlation between the two records through time. These analyses showed that the correlation coefficient between δ13Cinorganic and δ13Corganic increased from the proximal location (Site 1005, r2 = 0·1), to the distal site (Site 1006, r2 = 0·63). The importance of platform-derived carbonate and organic material at the proximal location, Site 1005, is reflected in the absence of a co-variation between inorganic and organic δ13C records, which exhibit no correlation on the platform itself. In contrast, the co-variance in δ13C values at the basinal location, Site 1006, is explained by a two-point mixing model, which demonstrates the importance of both pelagic and platform-derived carbonate and organic carbon in generating the positive correlation between the organic and inorganic δ13C values; this results in a correlation between δ13Cinorganic and δ13Corganic records at Site 1006 that is unrelated to global carbon cycling. Such data question the applicability of using δ13Corganic values to support the ability of δ13Cinorganic values to record global carbon cycling in carbonates recovered from environments where multiple sources of carbonate and organic carbon contribute to the bulk δ13C signal.