Interpretation of Miocene and Pliocene global sea level history requires juggling a "chicken and egg" approach to conflicting proxies of benthic oxygen isotope stratigraphy and paleotemperature estimates using Mg/Ca ratios vs. more direct estimates of ice volume, global sequence stratigraphy of sea level change (Hag et al, curve), and regional tectonics. Different approaches produce incongruent records, especially for particular time periods. For example, the relative early Miocene warmth suggested by less negative benthic oxygen isotope records (e.g., Zachos e ta.,2001) is not matched by paleoenvironmental evidence from the Antarctic margin (e.g., Cape Roberts Project). On either side of the early Miocene Mi-1 event, sequence stratigraphic sea level curves and isotopically-based estimates of ice volume and deep sea temperatures are out of phase. In contrast, during parts of the Neogene, eustatic and isotopic records are consistent, especially during the Mid-Miocene climatic optimum. On the other hand, the gradual isotopic decrease (cooling) from 23 to 17 Ma (Billups and Schrag, 2002) is not evident in the eustatic sea level curve. Pliocene estimates of ice volume based on combined del 18O and Mg/Ca data does not leave much room for the growth of Northern Hemisphere ice sheets around 3 My, consistent with records for high latitude Northern Hemisphere warmth. Improved models of lithospheric flexure and local tectonics are needed to compare and calibrate evidence of high elevation paleoshorelines with best estimates of global ice volume.