Digital libraries are increasingly common, being developed by government agencies to disseminate and preserve the documentation produced by its employees. This proposes a challenge in describing this type of documents, dealing official aspects in tools that are originally designed for bibliographic and scientific documents. In this sense, our objective is to verify how digital libraries, linked to the executive, legislative and judiciary Brazilian powers, are describing its documents collections. A study with descriptive and qualitative characteristics reveals the great adoption of DSpace software for creating these digital libraries and Dublin Core to describe the documents, showing DSpace and metadata schema adaptability for nonacademic document types. Thus, one contributes to the discussion on the use of Dublin Core to describe various types of documents on the Internet.