For the standards SRM(NIST) 987, EN-1, and E&A, we present compilations of 87Sr/86Sr from the literature, and 125 new measurements of 87Sr/86Sr, to arrive at recommend values of 0.710 248, 0.709 172, and 0.708 021 respectively, for normalising 87Sr/86Sr data for use in Sr-isotope stratigraphy. Using these standards, interlaboratory comparison of 87Sr/86Sr data is, nevertheless, impaired by residual bias of up to 0.000 072 that is not removed by normalisation of 87Sr/86Sr data to accepted values for standards. We also provide a new reference curve of 87Sr/86Sr against time for the period 11 to 20 Ma that is based on high-precision (± 0.000 003, 2 s.e.) analysis of 87Sr/86Sr in microfossil calcite from the ODP Sites 1146 and 1264, and IODP Sites U1337 and U1338. Ages for Site U1338 are calibrated by bio-magnetostratigraphy, whilst the others are calibrated by astrochronology, within which anomalies of up to 1.0 myrs are noted. With literature data, we extend the curve to 0 and 32 Ma. We show that the previous 87Sr/86Sr reference curve for the interval 0 – 7 Ma, that of Farrell et al. (1995) for Site 758 in the southern Bay of Bengal, is compromised by riverine influences from the Ganges/Brahmaputra rivers: differences between that curve and our new curve are interpreted to be a proxy for the intensity of the Indian summer monsoon. Finally, for Site 1120, a new age-model is presented, based on 87Sr/86Sr, in order to test the biostratigraphic age-model for the site, which is found to be good. This work shows that integrated approaches to improving the chronological framework of Cenozoic successions should include Sr-isotope stratigraphy.