Several lines of evidence, both previously published and unpublished, are used to illustrate deep-sea connections that have been either constant or fluctuated throughout the Quaternary in the northwest Atlantic. The best record is from a 100m Quaternary section the Central Labrador Sea (CLS, 3700m) obtained during ODP Leg 105, an HPC core over 100m long at site 647B. This site was sampled at relatively high resolution and shows periodic intrusions of Antarctic Bottom Water (AABW) as indicated by the strong presence of the foraminifer Nuttalloides umbonifera during interglacials from the present to approximately 1Ma and co-dominating with several other species below the1Ma mark. In the same core the Arctic Bottom Water (ABW) indicator species, Stetsonia arctica, is present in most of the interglacials between Isotope Stage 5 and 32 ( approximately 1 Ma) but occurring in different parts of the interglacials than N. umbonifera, except not in the Holocene. Significantly, N. umbonifera is present in the Holocene. These two species indicate strong connections between the Arctic and Antarctic to the Labrador Sea during interglacials. During glacials the Labrador Sea is often occupied by local glacial bottom water (GBW) indicated by the presence of Uvigerina peregrina. In contrast, the Arctic Ocean, as shown in previously published work, does not indicate (in the central Arctic) much influence from the Atlantic until about 400ka, when typical N. Atlantic bottom water species begin to invade and rapidly replace one of the endemic Arctic species, Bolivina arctica, but not S. arctica. This may be the reason that B. arctica does not seem to occur in the N. Atlantic with S. arctica. Previous work from the Norwegian Sea, both from surface samples and cores, do not indicate the presence of S. arctica or N. umbonifera either now or at any time in the Quaternary in the Norwegian Sea. The large sieve used (150 microns) explains the absence of S. arctica, but N. umbonifera was not present. Hence although ABW penetrates into the N. Atlantic throughout the Quaternary, Atlantic water does not enter the Arctic until about 400Ka. AABW appears to be present throughout the Quaternary but has not penetrated to the Norwegian or Arctic Oceans. There appears to be no deepwater connection to Labrador Sea from Baffin Bay, as had been suggested prior to Leg 105.