In recent years, increased attention has turned towards the role of coastlines in facilitating the global Pleistocene dispersal of humans (e.g. hypothesized coastal routes 'out of Africa' and into the Americas). However, research is hindered by a current lack of archaeological data due to submergence by sea-level rise, and the possibility that undiscovered sites have been reworked or destroyed by marine processes. Further, a current focus on the site as a unit of analysis means that the nature of the coast between sites and its influence on movement is less well understood. This is suggested by the fact that common assumptions describe the coast as stable and homogenous, a view that contrasts with other disciplines (e.g. geomorphology). This thesis outlines an approach that examines coastal environmental data, from a variety of disciplines, at large spatio-temporal scales. This circumvents the problem posed by current lack of data, the focus on individual sites and also provides a timely means of integrating increasingly high-resolution palaeo-environmental and palaeoceanographic data with archaeology. By first considering the data on a global level aimed at identifying generic patterns of spatio-temporal change in the coastal system, it was demonstrated that assumptions of stability and homogeneity are invalid for the Pleistocene and the process of colonization involved more than just passive funnelling along an undifferentiated corridor. More refined analysis was undertaken using a continental-scale case study (colonization of the Americas). This demonstrated that conclusions above are valid within a more restricted spatio-temporal window and provided hypotheses as to why and how people used the coast in this instance. Finally, a regional-scale example (re-colonization of NW Europe) provided further examples of coastal change and demonstrated that oceanographic fluctuations potentially exerted a strong influence on demographic changes seen in the archaeological record.