This paper describes an intelligent decision support system for information on petroleum resources and use currently being designed by the Information Methodology Research Project as the first step in the development of a comprehensive intelligent information system for dealing with energy resources in the United States. The system draws on artificial intelligence technology both in the design of the reasoning and representation components and in the computer language. The computer language supports a powerful declarative semantics based on frame hierarchies. The design supports desirable user features such as substitution for incomplete data, data validation, caveat warnings, and simple noticing. A model of oil flow in the United States forms the semantic basis for the system. A reasoning component within this representation provides answers to queries, a capacity for analyzing scenarios, and complex monitoring of the database for developing trends. References are listed. (Author/RAA)