As business rules are very common in business information systems, it is fundamental to analyze geographic rules in order to include them into geographic knowledge systems. The goal of this paper is, starting from examples taken overall from urban and environmental planning, to examine geographic rule semantics, and in particular to model space by combining logic and computational geometry. They can come from laws, physical laws, socio-economic considerations, from best practices, or from data mining. By defining new concepts such as metarules, superseded rules and jurisdictions, a conceptual model is proposed