This dataset contains single-shell stable isotope data (d18O and d13C), shell weights, ages, and preservation states of Cibicidoides shells spanning MIS 9-10 from ODP Site 1089 north of Agulhas Ridge (-40.936°N, 9.894°E, 4620 m water depth). Sediment samples were processed following standard procedures. From each >63µm sediment sample, all specimens of the genera Cibicidoides and Uvigerina were picked from the >150 µm size fraction. Using a binocular microscope, each individual Cibicidoides and Uvigerina specimen was critically evaluated for preservation quality. Details of the Foraminiferal Preservation Index (FPI) can be found in the corresponding publication. In short, an initial score (IS) of 5 is applied to an ideal specimen with clearly defined whorls, sutures, pores, and other surface features. A shell with an IS score of 3 usually contains enlarged pores and “smoothed” looking surface features, including sutures and surface ornamentation. An IS score of 4 is assigned to specimens with more subtly smoothed features, and/or slightly enlarged pores. From the IS3-5 value for a given specimen, deductions of 1-2 points are applied to generate its final score (FS). A deduction of 1 point is applied for either a missing or fractured chamber, and/or evidence of slight dissolution. A deduction of 2 points is applied for extensive missing or fractured chambers, and/or evidence of extensive dissolution. For example, an assigned FS1 value for a specimen could be the result of highly smoothed surface features, with some combination of extensive dissolution pockmarks and/or one to several fractured/missing chambers (i.e. IS3 – 2 = FS1). The letter “A” assigned to the final score indicates that the appearance of the shell is glassy, “B” is assigned to pseudo-glassy shells, and “C” to frosty specimens. Specimens of varying preservation quality were purposely chosen for single-shell stable isotope analyses. Each specimen was weighed, and shells featuring narrow mass ranges were selected for individual analytical runs of d18O and d13C which allows for the optimization of instrument tuning. Single Cibicidoides shells were analyzed on a Thermo Delta V Kiel IV dual inlet mass spectrometer at the Lamont-Doherty Earth Observatory, and were corrected relative to the NBS-19 calcite standard. Average precision on these analyses was better than 0.08‰ for δ18O and better than 0.06‰ for δ13C. The age model is based on the published stable isotope record (Hodell et al. 2001) tuned to the deep sea benthic d18O stack (Lisiecki and Raymo, 2005; Ahn et al., 2017).