Abstract : The aim of the SRCMF project is to mark up syntactic structures in the texts of two major old French corpora, the Base de Français Médiéval and the Nouveau Corpus d'Amsterdam. The annotation model is dependency-based. At first, the texts are marked up manually. Later, this annotation is used as "gold standard" to train automatic parsers. The corpus can currently be searched using TigerSearch software. Project documentation will be published on the web as soon as it gets stable stable, and access to the corpus will be provided to researchers upon motivated request.