AbstractAs the volume of digital data is growing exponentially and computational methods are advancing rapidly, network analysis is an increasingly important analytical tool to understand social life. This paper revisits the rich history of network analysis in geography and uses insights from that history to review contemporary computational social science. Based on that analysis, we synthesize the distinctive qualities of what we term geographical network analysis. Geographical network analysis presumes that networks are situated, construed through meaning, and reflect power relations. Instead of pursuing parsimonious explanations or universal theories, geographical network analysis strives to understand how uneven networks develop across space and within place through a constant back and forth between abstraction and contextualization. Drawing on the articles in this special issue, this paper illustrates how geographical network analysis can be put to work.