In July 1947, the British government of India convened two boundary commissions for the purpose of partitioning the colony into two independent nation states: India and Pakistan. Despite the volume of scholarly work on partition, little attention has been paid to its geographies. This paper addresses this lacuna in two ways: first, by inserting a geographical account into wider historiographical narratives of partition through cartographic-visual analysis and the application of geographical literature. Second, by bringing some of the themes and theoretical contributions from literature on partition into a geographical framework by highlighting the ways in which technical geographical terminology and boundary making practices were used for the political purpose of claiming territory. The paper pursues both of these aims through a biographical approach, examining in detail the unexamined papers of the British geographer Oskar Spate and his minor yet revealing involvement in the Punjab Boundary Commission hearings of July 1947. The paper argues that a territorial divide was unable to resolve the tensions that lay at the heart of the process of partition. This was not simply because geographical expertise was disregarded, but because geographical knowledge and analysis could not provide an adequate solution to the problem of overlapping yet ideologically distinct imagined territorial homelands. The paper thus points to the varied and ambiguous role that maps and mapping played in post-war sites and projects of decolonisation.