AbstractThe last five years has seen a considerable growth in the application of graph and network theory to “real‐world” networks. Many of these networks are “spatial” in that their form and dynamics are limited by the distance between nodes in Euclidian space. This paper reviews some of the advances in the analysis of such real‐world networks, highlights the continuing difficulties involved, and points to areas that need further work before a network analysis can become a standard field of application for those wishing to understand and predict real‐world systems. © 2010 Wiley Periodicals, Inc. Complexity 16: 11–15, 2010