Description: Data consist of a continuous millennially-resolved record of oxgyen and carbon stable isotope composition of planktonic and benthic foraminifera at IODP Site U1385 and piston core MD01-2444 from the southwestern Iberian margin for the last 1.5 million years. Elemental ratios (Ca/Ti and Zr/Sr) are also reported from the composite section of Site U1385 and core MD01-2444. Oxygen and carbon isotope measurements of planktic and benthic foraminifera were made at an average temporal resolution of ~200 years for the last 1.45 million years. For planktic foraminifera, the surface-dwelling species Globigerina bulloides from the 250 - 350 um size fraction was used. Benthic foraminifer data consists mostly of Cibicidoides wuellerstorfi and occasionally other species of Cibicidoides were measured from the 〉212 um size fraction. In samples where specimens of Cibicidoides spp. were absent, we used d18O of Uvigerina peregrina or Globobulimina affinis. All δ18O values for each species were corrected to Uvigerina using the offsets suggested by Shackleton et al. (2000) -- i.e., +0.64 for Cibicidoides and -0.3 for G. affinis. Stable isotope measurements were made in the Godwin Laboratory for Palaeoclimate research at the University of Cambridge. Foraminifer tests were crushed and soaked in a solution of 1% hydrogen per-oxide for 30 min in individual vials. Acetone was added and the samples placed in an ultra-sonic bath for 10 s, after which the liquid was carefully decanted to remove any contaminants. The samples were dried in an oven at 50 °C overnight. Isotopic analyses of the samples were per-formed using a VG SIRA mass spectrometer with a Multicarb system for samples with a mass exceeding 80 μg. Analytical precision is estimated to be ±0.08‰ for both δ18O and δ13C. For smaller samples (〈80 μg), measurements were performed on a Thermo Finnigan MAT253 mass spectrometer fitted with a Kiel device. Analytical precision is estimated to be ±0.08‰ for δ18O and ±0.06‰ for δ13C, respectively. All results are reported relative to VPDB. Elemental ratios (Ca/Ti and Zr/Sr) were obtained using an Avaatech X-ray fluorescence (XRF) core scanner at 1-cm spatial resolution. Cores from Holes U1385A and U1385B were measured at the Royal Netherlands Institute for Sea Research (NIOZ) and Holes U1385D and U1385E and MD01-2444 were measured at the Godwin Laboratory for Palaeoclimate Reseach at the University of Cambridge. The core surface was carefully scraped cleaned and covered with a 4-μm thin SPEXCertiPrep Ultralene foil to avoid contamination and reduce desiccation. All XRF scanners irradiated a surface of 10-mm high by 12-mm wide every 1 cm with identical instrument settings used at both NIOZ and Cambridge. Cores were scanned at 3 voltages: 10kv (750 μamps, 10 sec count time with no filter); 30 kv (500 μamps, 10 sec count time with Pd thin filter); and 50 kv (1000 μramps, 60 sec count time with Cu vilter). Element intensities were obtained by post-processing of the XRF spectra using the Canberra WinAxil software with standard soft-ware settings and spectrum-fit models. The purpose of the analyses were to produce records of millennial-scale climate variability (MCV) during the last 1.45 million years. Questions to be addressed with the data included: How common was MCV during older glacial periods of the Pleistocene? Did the nature (intensity, duration, pacing) of MCV change with orbital configuration or climate background state (ice volume, sea-level, ice sheet height)? What is the relationship between MCV and longer-term, orbitally-driven glacial-interglacial cycles – how do they interact? How did MCV change across the Middle Pleistocene Transition (MPT) when ice sheets grew larger in size and the amplitude of glacial-interglacial cycles increased? Was the thermal bipolar seesaw mechanism active during older glacial periods of the Pleistocene? What role did millennial variability play in atmospheric CO2 variations or vice-versa?