A system and method for preserving unrepresented data in a document exchanged between applications. Applications handling the document, or computing devices executing the applications may have different capabilities with respect to a particular feature of the document such as certain formatting properties, footnotes, graphical attributes, and the like. Data associated with such features, unrepresented data, is preserved by employing a tree structure for the document and flagging nodes associated with unrepresented data. Flagged nodes are ignored, but preserved, when the document is handled by an application or computing device with different capabilities. When the document is handled again by an application or computing device that is capable of processing those features, the unrepresented data is made available along with any modifications made to the document during the previous handling.