This article proposes a method for document/speech align- ment based on explicit references made in speech to documents and parts of documents, in the context of multimodal meetings. The motivation and the main components of the method are —rst described brie°y. Then, the article focuses on the two main stages of dialogue processing: the detec- tion of the expression referring to documents in transcribed speech, and the recognition of the documents and document elements that they refer to. The detailed evaluation of the implemented modules, —rst separately and then in a pipeline, shows that results are well above the baseline, and that the various features of the proposed algorithms are all relevant. The integration of this document/speech alignment technique with other ones is —nally discussed.