The paper introduces an ongoing project for the development of a parallel treebank for Italian, English and French annotated in the pure dependency format of the Turin University Treebank, i.e. Parallel–TUT. We hypothesize that the major features of this annotation format can be of some help in addressing the typical issues related to parallel corpora, e.g. alignment at various levels. Therefore, benefitting from the tools previously used for TUT, we applied the TUT format to a multilingual sample set of sentences from the JRC-