The success of early hydrogen vehicle markets will be contingent on the resolution of a chicken and egg problem: consumers will demand an adequate refueling network before purchasing hydrogen vehicles, and investors will require sizeable market demand before building a refueling network. In this research project we attempt to provide insights into the issue of adequate refueling. Under the assumption that consumers consider the current gasoline network adequate, we characterize existing spatial relationships and density (gallons dispensed) distributions. The current refueling network has an emergent structure; it is the result of over a century of changing demands and technologies. For this reason it is difficult to know the complete details of the network: no one knows them in full. Our analysis reveals a structure that is consistent across urban areas of all types. Our research is based on our analysis of gasoline consumption data for more than 50 U.S. urban areas. A log linear relationship between urban population and station count was found. In addition, an inverse correlation was found between population density and station density. Using statistical methods, we estimate the number of urban stations in the United States and compare our results with alternate data sources. Algorithmic clustering of the stations reveals remarkable similarities between station networks, regardless of city area, population or density. The top five percentile of stations sell five to eight times the gasoline of an average station. The top quartile of stations sells at least 150% of the average. The bottom quartile of stations sell less than a third of the average station. Clustering stations within one and two miles-a range that approximates adequate refueling availability-reduces an urban area's station count by more than two thirds. These results provide insights into the types of investments and stakeholder decisions required to establish early urban refueling station networks.