This article responds to a recent call from the journal’s editors to diversify the subjects of historical geography. It surveys Historical GIS (HGIS) and colonial cartography scholarship to offer a critique of existing GIS tools for the study of precolonial nonwestern histories. These tools, in the ways in which they turn Indigenous knowledge into computer-processed data, repeat the work of the colonial state. The article applies concepts from decolonial and design theory to address the problem, in particular Ngũgĩ wa Thiong'o’s ideas about the decolonization of knowledge. If we think of cartography as as a language to communicate spatial information, then a decolonial map must be based in Indigenous mapmaking. To articulate what this might look like, the article explores Karl Weule’s work on Indigenous mapmaking in early-20th century mainland Tanzania (then the colony of German East Africa). It demonstrates that the most common means of applying GIS to historical maps distort the mapmakers’ original intentions and proposes an alternative design for a pluralistic decolonial map, based in Indigenous conceptions of space rather than just incorporating Indigenous knowledge.