attractiveness potential than traditional administrative-based destinations. With these premises we proposed a method to delineate ‘tourism zones’. We have then implemented the method in the Pyrenees, which exhibits unique charac- teristics of mountain regions (Godde et al., 2000). We found nine tourism zones which differ from the current tourism destinations. They are more uniform in size than the administrative regions; all of them are cross-border either interregionally, internationally or both, and they are very similar to historical regions, thus sharing common cultural arte- facts like language, despite being divided by current borders. The results also show a high and significant correlation between the intensity and specificity of attractions and, in the case of tourism zones, the significant correlation between these and the number of accommodation units. Thus, the results indicate that unique first-class attractions go together with high numbers of lower level attractions nearby and that all together they foster the growth of accommodation units. However, the correlation between intensity/specificity of attractions and accommodation units does not occur in the case of administrative regions and, therefore, tourism zones are more effec- tive than administrative regions in representing the actual tourism structuration of space. The results also suggest that the more the tourism zones differ from administrative regions the higher their attractiveness, indicating that around administrative borders the accumulation and relevance of attractions are higher. Four types of mountain tourism zones were finally identified, ranging from high to low tourism intensity. Some of the most tourism-intensive types were located in the cen- tral and peripheral coastal regions of the mountain range and have the highest interna- tional cross-border intensity. These results can inspire further research on the identification of tourism zones in other regions of the world and in all types of geographi- cal settings (i.e. mountain areas, rural areas, urban areas or cross-border areas) and whether the higher relevance of tourism zones over administrative regions is also mani- fested in them. The results also have major implications for the management of destinations. On one hand, current destinations should be aware of the way they belong to or are divided into tourism zones. This awareness must be the basis for neighbouring destinations to strategi- cally initiate cross-border collaboration agreements, integrate their tourism policies and even create new tourism destination brands, which can be more meaningful for visitors and more satisfying for the travel patterns they seek. We can already see some moves in this direction in the coastal area of Basque Pyrenees and in the area of Cerdanya. The valley of Cerdanya is divided into two administrative regions, one belonging to France and the other to Spain. It is a historical, cultural region, with a common language and shared traditions and certain tourism functionalities. For example, accommodation, mostly placed on the Spanish side, also serves the ski resorts on the French side. In fact, cross-border initiatives in the Cerdanya have led to the development of cross-border man- agement and promotion actions, such as the creation of cross-border promotional mate- rial. These examples suggest interesting perspectives for research in the field of tourism policy and planning. The creation of new tourism zones in cross-border regions entails important political challenges. Tourism stakeholders face additional impediments to cross-border development and their role seems to be crucial in the performance of cross- border collaboration strategies. In this sense, some stakeholders may play a more subser- vient role, while others play a more dominant role. Policy integration processes are needed between different regions sharing tourism zones, and marketing strategies need to be amended to address specific market niches, which may vary, to a certain extent, from previous conceptions of tourism space. For example, newly found tourism zones may