Application of volcanic proxies to paleoenvironmental studies in Iceland has generally lagged behind the improvements of undisputable geochronologies. This is particularly frustrating because a link between volcanism and deglaciation in Iceland is an appealing hypothesis and questions about the extent of the ice cap during Icelandic glaciations remain pending (Einarsson et al., 1988; Geirsdttir et al., 1997; Ingolfsson et al., 1997; Hoppe, 1982). The postglacial eruptive history of Iceland after the last deglaciation is now well documented (Licciardi et al., 2007; Sinton et al., 2005, Saemundsson, 1991). One of its main characteristics is an increase in the volcanic eruption rate, which coincides with the end of the last ice age (11 ka BP; Slater et al., 1998; Maclennan et al., 2002; Sigvaldason et al., 1992; Hardarson et al., 1991; Jull et al., 1996). However, the pattern of volcanism for more evolved rocks, such as rhyolites, comprising 10-12% of outcrop on Iceland, is less well established. In fact, the opposite trend may be apparent, or at least sub-glacial rhyolite eruptions are frequent during interglacial periods (Flude et al., 2008). This relationship between deglaciation and volcanism has been convincingly demonstrated in Iceland only for the last deglaciation (Maclennan et al., 2004; Sinton et al., 2005; Licciardi et al., 2007, McGarvie et al., 2006, McGarvie et al., 2007). Therefore, we consider that it is important to explore this relationship through older periods. The purpose of this study is then to evaluate the potential of the unspiked K-Ar dating method in dating Quaternary Icelandic volcanics to contribute to this debate. The unspiked K-Ar technique applied here is perfectly adapted to these purposes because it is an accurate and precise dating tool which has proven to be efficient and suitable to date young Quaternary lavas. To test the relevance of our approach, we collected and dated sub-aerial lavas contemporaneous with warm periods; hyaloclastites and lavas formed in sub-glacial environments.