In this article I outline the ideas behind the Open Archives Initiative metadata harvesting protocol (OAIMH), and attempt to clarify some common misconceptions. I then consider how the OAIMH protocol can be used to expose and harvest metadata. Perl code examples are given as practical illustration.