In this paper a visual approach to querying in large spatial databases is presented. A diagrammatic technique utilising a data flow metaphor is used to express different kinds of spatial and non-spatial constraints. Basic filters are designed to represent the various types of queries in such systems. Icons for different types of spatial relations are used to denote the filters. Different granularities of the relations are presented in a hierarchical fashion when selecting the spatial constraints. The language constructs are presented in detail and examples are used to demonstrate the expressiveness of the approach in representing different kinds of queries, including spatial joins and composite spatial queries.