In the history of mountain villages of Japan, it was during the medieval period when mountain villages advanced the frontiers of land use and started actively territorializing mountain space. But the status of medieval mountain studies is restricted for lack of data over various regions. Thus we have to find a new way of reading materials. My purpose is to illuminate physical features in the context of medieval development of mountain villages.What implications did mountain landforms have for medieval utilization or territori-alization? We have tended to think of ridges and valleys as major bases of space recognition, land use, and demarcation in mountainous areas. But Chiba (1986) suspects that mountainous topography was hardly the basis of village boundaries in the Edo period (18th century) for mountain villages based on a subsistence economy of shifting cultivation. Were not medieval mountaineers very conscious of mountain texture? Did the spatial unit of catchment area have little significance for land use?I investigate the medieval documents associated with the demarcation between myos (lower units of a manor, which were originally used to collect taxes) in Makinoyama situated in the eastern Shikoku mountainous region (see Figure 1). This is also the field for Chiba's research mentioned above. Every myo in Makinoyama had a territory which contained mountainous space. The documents range from the 13th century through the 15th century, and myo boundaries were indicated in some of them.Some of the medieval myo's territories were inherited until the Edo period, when both accords and discords between the water system and political borders, as well as the territory enclave, were observed (Figure 2). However, most boundaries was corresponded to ridge or river in each place, and territory consisted of several small drainage units. It is presumed that most of the enclaves were formed through division and transfer of several myo territories in the 15th and 16th centuries (compare Figure 4 with Figure 3).In the examination of the documents, special attention is devoted to following points of view associated with land recognition and land use. First point concerns demacators' understanding of physical and social dividing lines. How did readers of the documents reconstitute such boundaries in their own spatial recognition? What understanding of mountain space could the demarcators utilize in the process of demarcating? Second, were ridges and valleys obstacles to land use for medieval mountaineers? How are possible answers to this question related to demarcation of myo territory?The finding obtained are summarized as follows: First, readers of the demarcation documents have to reconstitute boundary line from description of physical borders and a variety of spots or landmarks (such as trees, crags, falls, river canyons, ponds, damp ground, hills and buildings). Medieval mountain people became aware of space and boundary based on the physical boundaries and the points within the mountain area in question.Second, in the 13th and 14th centuries, several myos contended over boundaries repeatedly. The disputes occurred due to expanding land use of shifting cultivation/slope agriculture and gathering. A detailed inspection of their locations indicates that all the disputes happened at lower elevation (less than 800m), which were located at a similar elevation to the settlements concerned and were accessible from them. This suggests that, with regard to such lower places, ridges and valleys did not disturb land use development, and some valleys were a units of land use and dispute. It is surmised that the boundaries were fixed through repeated disputes at various places.