After an extensive study of the metadata policy of each of its content partners, the EuDML project evaluated many different strategies and existing schemas that could store every detail faithfully, and yet reserve room for the enhancements foreseen in the project's work plan. The framework provided by the so-called NLM Journal Archiving and Interchange Tag Suite was selected as best readily available approximation of our needs. Some modifications of it have been endorsed by the project, defining the first version of our interchange schema for heavily mathbased content.