In recent years, economic geographers have seized on the concepts of 'path dependence' and 'lock-in' as key ingredients in constructing an evolutionary approach to their subject. However, they have tended in to invoke these notions without a proper examination of the ongoing discussion and debate devoted to them within evolutionary economics and elsewhere. Our aim in this paper, therefore, is, first, to highlight some of the unresolved issues surround these concepts, and, second, to explore their usefulness for understanding the regional economic evolution. We argue that in many important aspects, path dependence and lock-in are place-dependent processes, and as such require geographical explanation. At the same time, there has been little discussion of regional path creation: te assumption has been that new technological-economic paths emerge at random or spontaneously across space, an assumption that we find too simplistic. This leads on to the key question as to why some regional economies become locked into development paths that lose dynamism, whilst other regional economies seem able to avoid this danger and in effect are able 'reinvent' themselves through successive new paths or phases of development. We conclude that whilst path dependence is an important feature of the economic landscape, the concept requires further elaboration if it is to function as a core concept in an evolutionary economic geography.