The many different proxy records from the European Project for Ice Coring in Antarctica (EPICA) Dome C ice core allow for the first time a comparison of nine glacial terminations in great detail. Despite the fact that all terminations cover the transition from a glacial maximum into an interglacial, there are large differences between single terminations. For some terminations, Antarctic temperature increased only moderately, while for others, the amplitude of change at the termination was much larger. For the different terminations, the rate of change in temperature is more similar than the magnitude or duration of change. These temperature changes were accompanied by vast changes in dust and sea salt deposition all over Antarctica. Here we investigate the phasing between a South American dust proxy (non-sea-salt calcium flux, nssCa2+), a sea ice proxy (sea salt sodium flux, ssNa+) and a proxy for Antarctic temperature (deuterium, δD). In particular, we look into whether a similar sequence of events applies to all terminations, despite their different characteristics. All proxies are derived from the EPICA Dome C ice core, resulting in a relative dating uncertainty between the proxies of less than 20 years. At the start of the terminations, the temperature (δD) increase and dust (nssCa2+ flux) decrease start synchronously. The sea ice proxy (ssNa+ flux), however, only changes once the temperature has reached a particular threshold, approximately 5°C below present day temperatures (corresponding to a δD value of 420‰). This reflects to a large extent the limited sensitivity of the sea ice proxy during very cold periods with large sea ice extent. At terminations where this threshold is not reached (TVI, TVIII), ssNa+ flux shows no changes. Above this threshold, the sea ice proxy is closely coupled to the Antarctic temperature, and interglacial levels are reached at the same time for both ssNa+ and δD. On the other hand, once another threshold at approximately 2°C below present day temperature is passed (corresponding to a δD value of 402‰), nssCa2+ flux has reached interglacial levels and does not change any more, despite further warming. This threshold behaviour most likely results from a combination of changes to the threshold friction velocity for dust entrainment and to the distribution of surface wind speeds in the dust source region.