A web-based tool for the identification of protein structures based on structural properties. SpeeDB queries are executed on all structures in the PDB at once, quickly enough for interactive use. SpeeDB includes standard queries based on published criteria for identifying various structures: disulphide bonds; catalytic triads; and aromatic-aromatic, sulphur-aromatic, cation-π, and ionic interactions. Users can also construct custom queries in the user interface without any programming. Results can be downloaded in a CSV format for further analysis with other tools.