Reconstructions on Antarctic ice cores revealed pronounced, millennial-scale variabilities in atmospheric CO2 over the past 800,000 years (e.g. Lüthi et al., 2008; Monnin et al., 2001; Petit et al., 1999; Raynaud et al., 2005; Siegenthaler et al., 2005). Despite these variabilities are known for several decades, the mechanisms, driving these patterns are still not fully resolved. As the ocean contains up to 60 times more carbon than the entire atmosphere, it is considered to be a major driver of the atmospheric CO2 levels (Broecker, 1982): Storing CO2 during glacials, releasing it during deglaciations. Because changes in the global thermohaline circulation are thought to operate on glacial/interglacial timescales, it has been suggested that during glacials, the deep ocean was separated from surface-waters and therefore from the atmosphere by enhanced stratification, resulting in the pronounced accumulation of CO2 and nutrients in the lower levels of the water column. These waters, isolated for millennia, then surfaced during interglacial periods and released their load of ancient CO2. This hypothesis is strongly supported by the record of atmospheric radiocarbon activities (Δ14C; Reimer et al., 2013) in which a telltale drop in Δ14C is shown during the interval of enhanced increase in atmospheric CO2. This drop cannot be explained by the atmospheric formation of 14C and is therefore indicative for the release of an old and hence 14C-depleted and CO2-enriched reservoir (e.g. the deep ocean) to the atmosphere. This release can therefore explain both records (atmospheric CO2 and Δ14C). Indeed, several records from the Atlantic and Pacific Oceans point to the existence of this carbon pool in glacial deep-waters below 2000 m. However, the spatial extent (vertical and lateral) and particularly the pathways of upwelled waters during the interglacial remain elusive and even contradictory. The aim of this thesis is to improve the knowledge of changes in South Pacific circulation and ventilation over different glacial to interglacial time-scales. The three manuscripts that form the backbone of this thesis are used to: 1) constrain the spatial extent and upwelling pathways of the glacial carbon pool in the South Pacific (0 – 30,000 years); 2) reconstruct boundary shifts between intermediate-waters and the underlying carbon pool in Circumpolar Deep Water over the last 350,000 years; 3) analyze changes in the South Pacific Gyre’s thermocline during the past 200,000 years. The focus of the first manuscript (Chapter 3) lies on the transition from the last glacial to the current interglacial. The Δ14C-reconstructions on a water mass transect of seven sediment cores from the New Zealand Margin and the East Pacific Rise identify a pool of radiocarbon depleted waters between ~2000 and ~4500 m in the glacial counterpart of Pacific Deep Water. The 14C-depletion of this body of water is up to five times higher than in the modern South Pacific and reaches extreme apparent ventilation ages of ~8000 years. Despite the first deep water rejuvenation begins as early as ~21,000 years B.P., the main signal of rejuvenation and outgassing parallels the rise in atmospheric CO2. The vertical extent of southwest Pacific Antarctic Intermediate Water (AAIW) over the last four glacial/interglacial cycles is analyzed in the second manuscript (Chapter 4). Stable isotope records (δ13C and δ18O) from epibenthic foraminifera of sediment cores bathed in AAIW and Upper Circumpolar Deep Water (UCDW) indicate a shoaling of AAIW during glacial periods. Further support for these findings arises from model reconstructions using the CCSM3-climate model. Throughout glacial maxima, pronounced input of freshwater by melting sea ice into the AAIW significantly increased its buoyancy and hampered its downward expansion. Hence, the upward displacement of the AAIW-UCDW boundary led to an expansion of the glacial carbon pool identified in Chapter 3. In the third manuscript (Chapter 5) the evolution of Southern Ocean Intermediate Waters (SOIWs) is analyzed, using the Mg/Ca paleothermometry on surface- and deep-dwelling species of planktic (Emiliani, 1991) foraminifera. The results suggest opposing glacial subsurface conditions during the LGM and MIS 6 with colder-than-Holocene conditions during the former and warmer-than-LGM conditions during the latter interval. Because of the importance of SOIWs for the ventilation of the South Pacific Gyre (SPG), the results of Chapter 5 reveal the relevance of Southern Ocean subsurface processes on the transfer of climatic signals from higher to lower latitudes via the SPG. Ultimately, this thesis contributes to the broader understanding of ventilation and circulation changes in the Pacific Sector of the Southern Ocean. The combination of various proxies reveals the highly dynamic processes that affect the Southern Ocean on glacial/interglacial timescales. The results do not only constrain the vertical extent of the glacial carbon pool for the first time, but they also facilitate its integration in the global context of glacial circulation. Furthermore, the reconstructions shown in this study might help to improve model simulations that are used to both, reconstruct and predict changes in the global climate.