Historical observations of the concentration of calcium carbonate in global deep sea sediments are compiled and compared with a new gridded field of seawater CO3=concentration to reveal regional variations in the calcite lysocline. The most obvious mode of variability of the calcite lysocline is the thickness of the lysocline (defined here as the difference in overlying water carbonate saturation, ΔCO3=, between high and low calcite sediments) with a thicker lysocline in the Atlantic than in the Pacific. I attribute this variation to differences in the delivery rate of terriginous material. A recent model for the lower glacial atmospheric pCO2proposed to change the relationship between the depth of the lysocline and the ΔCO3=of the water column by changing the rain rate ratio of organic carbon to calcite production (the “rain ratio model”: Archer and Maier‐Reimer, 1994). I search the data set for analogs to the proposed glacial world, by looking for a link between the regional climate at the sea surface and the depth of the lysocline below. The ΔCO3=at the carbonate compensation depth (CCD) in the tropics appears to be 10–20 μmol kg−1ΔCO3=more undersaturated than in high latitudes, but this is smaller than the ∼40 μmol kg−1shift required by the model. In addition, the general resemblance of the glacial lysocline to the present day requires that the proposed shift in ΔCO3=at the CCD be globally uniform rather than locally variable, as climate forcing would probably generate. I conclude that the rain ratio model would probably require some globally uniform perturbation during glacial time, such as a change in ocean Si content, if it is to explain the entire pCO2decrease observed in the glacial atmosphere. Finally, I grid the sedimentary data to estimate that the inventory CaCO3which is available to neutralize fossil fuel CO2is approximately 1600 Gt carbon, a quantity which may be exceeded by fossil fuel release in the ne