Abstract
The Selective Harvester provides a flexible and customisable mechanism to select and re-use metadata records from open access repositories. The open-source Java-based application developed at the University Library, University of New South Wales (UNSW) integrates an existing OAI-PMH harvesting tool (jOAI), with an application which filters and ingests selected records into a Fedora repository. This model has applications in scholarly communication and eResearch services, especially in relation to populating subject-based repositories.