This work addresses problems associated with data exchange and data representation in the computational chemistry and physics communities. Recent computational developments, such as Condor and the Grid, have paved the way for new kinds of simulations that demand more rigorous data handling. To this end, the paper discusses the use of XML and the Chemical Markup Language (CML) in theoretical chemistry and physics. Extensions to the core CML language, known as CMLComp, are also discussed. However, the majority of atomic scale simulation software is written in Fortran. Fortran's lack of XML support represents a potential barrier to the adoption of CML in these fields. This has prompted the authors to develop XML and CML processing tools for Fortran, including native SAX and DOM implementations, as well as libraries for generating well formed XML and CML. These libraries have been used to extend existing simulation packages to work with the CML and CMLComp languages. Finally, we give a practical example that highlights how these XML aware applications can be effectively used as workflow components in complex chemical and physical simulations.