This paper addresses a data structure specification for route directions that incorporates essential aspects of cognitive information processing. Specifically, we characterize levels of granularity in route directions as the result of the hierarchical organization of urban spatial knowledge. We discuss changes of granularity in route directions that result from combining elementary route information into higher-order elements (so called spatial chunking). We provide a framework that captures the pertinent aspects of spatial chunking. The framework is based on established principles used—from a cognitive perspective—for changing the granularity in route directions. The data structure we specify based on this framework allows us to bridge the gap between results from behavioral cognitive science studies and requirements of information systems. We discuss the theoretical underpinning of the core elements of the data structure and provide examples for its application.