Throughout the Late Cretaceous to Eocene, sedimentation in gradually subsiding basins on the passive eastern margin of the micro-continent of Zealandia recorded climatic and paleoceanographic changes in a greenhouse world. One such fundamental change in Southern Ocean circulation is hypothesised to be recorded in a regionally extensive unconformity surface and short-lived lithofacies changes contained within Late Paleocene to Early Eocene sedimentary successions at key sections throughout New Zealand, and investigated here on Campbell Island and in southeastern Marlborough. On Campbell Island, this oceanographic event is represented by an unconformity between the Late Cretaceous to Late Paleocene Garden Cove Formation and the Early Eocene to Oligocene Tucker Cove Limestone. This unconformity signifies a major lithofacies change from Garden Cove Formation which consists of siliceous mudstone containing fine sand to coarse silt sized siliciclastic grains, pelletal glaucony grains and rare quartz pebbles, to a nannofossil and foraminiferal limestone containing little to no siliciclastic grains comprising the Tucker Cove Limestone. Geochemically this lithofacies change is characterised by a dramatic decrease in terrigenous supply and a shift from siliceous to calcareous productivity, along with a significant concentration of Zr and rare earth elements. Lithofacies at this site are inferred to record possible episodes of ice rafting and eventual unconformity formation by invigorated intermediate depth ocean currents which resulted in winnowing of seafloor sediments and concentration of heavy minerals. At the distal Mead Stream site in southeastern Marlborough, deposition of bio-siliceous sediments of the Mead Hill Formation and Amuri Limestone was locally disrupted by deposition of the Waipawa Formation, the lateral equivalent of an important hydrocarbon source rock identified in several of New Zealand‟s sedimentary basins. In outcrop, the Waipawa Formation at Mead Stream is characterised by a very distinctive rusty‟ brown fissile appearance, while in thin section, though radiolarians and sponge spicules are common, the overall fine grained nature of the unit makes identification of other components difficult. Geochemical proxies show a significant increase in terrigenous supply in the Waipawa Formation, along with an increase in siliceous productivity concomitant with a decrease in oxygenation at the site. Lithofacies changes through the Late Paleocene at Mead Stream suggest the site lay under a zone of upwelling which resulted in an increase in siliceous productivity during the Late Paleocene. At the more proximal sites of Muzzle Stream and Kaikoura wharf in southeastern Marlborough, Mead Hill Formation and Amuri Limestone are separated by an unconformity, overlain by Teredo Limestone. The Teredo Limestone is considered to be a lateral equivalent of the Waipawa Formation, but both the base and top of the Teredo Limestone are timetransgressive. This means that at Muzzle Stream the unit is contemporaneous with the Waipawa Formation (Late Paleocene), while at Kaikoura wharf the unit is entirely Early Eocene in age. At these sites, the Teredo Limestone Member of the Amuri Limestone is a calcareous greensand sometimes containing phosphatised limestone clasts and sharks teeth. In thin sections, the unit consists of well sorted, fine to very fine sand sized siliciclastic grains and fine sand sized pelletal and vermicular glaucony set in a calcareous matrix that shows evidence of secondary silicification. Unconformity formation and the subsequent deposition of the overlying Teredo Limestone record a period of invigorated intermediate depth ocean currents that resulted in the transport of siliciclastic grains and glaucony to these bathyal sites. This interpretation is supported by a palinspastic map of the Teredo Limestone that suggests the unit was deposited under different conditions than those responsible for the deposition of the bounding Mead Hill Formation and Amuri Limestone. This map also suggests the Teredo Limestone was deposited as a skin drift‟, here named the Clarence Drift, possibly under the influence of contour currents. Based on similarities between unconformities and lithofacies changes in Late Paleocene to Early Eocene sedimentary sections and an earlier, well documented event at the Cretaceous/Tertiary boundary in southeastern Marlborough, evidence for a period of enhanced siliceous productivity, invigorated ocean currents and possible episodes of ice rafting is suggested to be consistent with a brief period of Antarctic ice sheet growth during a phase of global cooling in the Late Paleocene. The possible identification of Antarctic ice sheets, ephemeral though they may have been, not only challenges long held beliefs that the Antarctic continent remained ice free during the early Paleogene greenhouse world but also questions the suggested mechanisms responsible for Antarctic ice sheet growth. The lack of ocean gateways in the Southern Ocean during this time effectively rules out thermal isolation of the Antarctic continent as a driver. Given that this period of ice sheet growth is contemporaneous with a documented period of enhanced global ocean productivity and terrestrial carbon accumulation and related draw down in atmospheric CO2, it is suggested this may represent the driver responsible for brief Antarctic glaciation during this period, though the postulated link requires further investigation.