The Protein Data Bank (PDB) has grown from an archive of seven entries in 1973 to a large and rapidly growing collection of more than 30 000 structures as of May 2005. The PDB now provides a key information resource for the structural biology community. A software framework has been developed that supports automation and scalability, and that can adapt to changes in data content and delivery technology, to permit future development. The mmCIF data dictionary and an accompanying extended dictionary of local data items provide the core ontology which is the basis for software development. The associated software tools exchange and validate data, create and load databases, translate data formats, and serve application program interfaces. They are made available to the scientific community through an open-source licence.