Climate change during glacial and deglacial periods was dominated by rapid variations on a millennial time-scale. A central, but as yet poorly understood element of this millennial-scale climate change behavior is an asynchronous phasing of climate between the poles [Blunier et al., 1998], known as the 'bi-polar seesaw'. It is likely that the variability in intensity of intermediate flow from high to low latitudes is closely linked to these climate changes in the polar regions (sensu [Stott et al., 2007]). Computer simulations also suggest that variations in intermediate water circulation are intricately coupled to climate change in the polar regions' [Broecker, 1998; Stocker and Johnsen, 2003; Sijp and England, 2006]. Whilst there is growing recognition of the importance of intermediate water circulation for the bi-polar seesaw, millennial-scale variations in intermediate water circulation in crucial parts of the world ocean are not known. We are interested in the history of southern sourced intermediate waters (Antarctic Intermediate Water - AAIW) in the Indian Ocean. The role of GAAIW in the bi-polar seesaw is of global relevance due to its large volume and associated energy storage capacity. In the Indian Ocean, results of a New Investigators project (by the PI) provided insights into the short-term surface- and intermediate water variability recorded in the Arabian Sea. Surface ocean records from core NIOP 905 from the Arabian Sea imply a direct in-phase relation of monsoonal change with the well-known Dansgaard-Oeschger events in the N-Atlantic [Ivanochko et al., 2005] in line with previous results [Altabet et al., 2002; Schulz et al., 1998]. Striking new results are that 1) surface ocean and intermediate depth changes occur out-of-phase and 2) the intermediate water variability shows a close relation with Antarctic climate history, hence hosting evidence for a bi-polar seesaw pattern. Whilst the surface ocean change reflects Northern Hemisphere climate change, the intermediate water record seems to be tied to that in the south. This implies a climate connection between the high and low latitudes most likely mediated by variations in glacial AAIW (GAAIW) strength, with GAAIW enhancements occurring during the HE's, opposite to the N-Atlantic record of overturning circulation (figure 1). A locally derived signal for core NIOP 905, however, cannot be ruled out at this stage and additional research in a more southern position is required. Here we propose to utilize high-quality sediment core 64PE304-80 from the Mozambique Channel. This core is particularly suited for paleoceanographic work because it combines a very high sedimentation rate (~25 cm/ka), excellent carbonate preservation and a crucial location, allowing to trace surface and intermediate water changes at the millennial-scale. We aim to find evidence of increased GAAIW in the Channel to verify the wider significance of the Arabian Sea results, and the surface ocean data will help unravel the Indo-Atlantic surface water exchange (passing through the Channel) via the Agulhas system, which is a crucial component of Earth's climate [Peeters et al., 2004].