Patents belong to the few types of public information that have a big impact on national and international economies. During the last years there have been great efforts in making patent data available electronically for the public via online services. But today's services provide heterogeneous data structures which makes automatic processing difficult. None of the services supports all user aspects, so that different services have to be combined. In this paper we present an ontology-based approach for representing patent metadata and describe a Patent Metadata Ontology (PMO) that models the major aspects of patent metadata. The advantage of our approach is to provide a homogeneous representation of patent metadata merged from different sources. It allows for identifying context and dependency information more easily than today's database-centric structures and interfaces.