Geographical clusters have become axiomatic in theories of creativity and innovation. Clustering is advantageous to firms to promote greater levels of innovation due to physical proximity and the networked relationships that are unleashed. Because of this, cluster theory has become a perennial feature in off-the-shelf urban development policy prescriptions rolled out across cities worldwide. The chapter seeks to provide a critique of this state of affairs, aimed at sobering the degree of enthusiasm to rush to clusters as the pre-eminent policy solution. First, the authors revisit key thinkers in economic geography who theorized agglomerating tendencies as a key dynamic within a framework that encompasses centripetal and centrifugal geographic forces. Second, they illustrate the roles that underlying geography and history play in shaping the possibilities for agglomeration and dispersal of innovation activities. Two empirical examples from the authors’ previous work on the geography of creative industries are briefly revisited to illustrate. The first is a creative industries mapping project that sought to empirically document economic activity in Darwin, Australia, a small, highly suburbanized and physically remote city not normally associated with big city innovation. The second example is of bootmaking in El Paso, Texas. Drawing inspiration from recent critical, and grounded, work in evolutionary economic geography the authors argue that theorization of the geography of innovation must remain attuned to deeper run, geographically-contingent and cumulative-causal processes that shape present possibilities.