It is now widely accepted that expansion of continental glaciers is one manifestation of profound paleoenvironmental change in latest Miocene time. The oxygen isotopic record in deep-sea cores is inferred to document brief pulses in glaciation, based on covariance of δ18O in benthic and planktonic foraminifera, which probably lowered latest Miocene sea level by a minimum of 60 m. Additional evidence of glacial activity in the oxygen isotope record is obscured by the small signal amplitude. Before further details of paleoenvironmental change can be investigated it is essential to have a high-resolution stratigraphy and chronology. This has not been achieved previously because most studies of the latest Miocene have had a sampling interval greater than the duration of events they seek to resolve. A high-resolution stable isotope record from the north Atlantic (Deep Sea Drilling Project (DSDP) 552) is correlated with a second record from the southwest Pacific (DSDP 588). Each core has magnetostratigraphy, but there are two possible interpretations of magnetic results in the Atlantic core. By assuming that brief events of 18O enrichment and 13C depletion in each core are synchronous it is possible to assign the most likely age model to the Atlantic core. Results of this exercise indicate that two δ18O maxima occurred at 4.8 and 5.2 Ma. If these events reflect continental glaciation, and if sea level was lowered enough to contribute to isolation and desiccation of the Mediterranean, then the Miocene/Pliocene boundary must be close to 4.8 Ma in age because Messinian deposits are latest Miocene in age. This interpretation is consistent with biostratigraphy at DSDP 552; magnetostratigraphy, biostratigraphy, and isotope stratigraphy at a site commonly referred to because of its proximity to the Mediterranean (DSDP 397, northwest African continental margin); and recent work on land in the Mediterranean region.