At Giessen university, a drug formulary comprising drug data and treatment guidelines is supplied to clinical users who can access the drug information by an index of drug substances and drug substance groups. The guideline itself is a textual description with related information such as drug substances and drug brand names. Since clinical users also want to access the information by drug names, we had to extract this information from the textual descriptions. The extraction however caused some effort. In order to not repeat this effort in the future, we used the extensible Markup Language (XML) to restructure the information sources. This paper describes our experiences with this kind of legacy to XML conversion and outlines a possible migration path towards the XML technology.