For a number of years, the SCEN has been setting up areas where vegetal techniques are used to stabilize river banks, in a number of cases for experimental purposes. A colloquium organized in Liège in May 1998 offered an opportunity for assessing the situation of these sites and compare it with the approaches followed in other countries. A debate was initiated on the principle of non-intervention, or the preferential use of vegetal techniques, in cases where nesting of given bird species isexpected. An inventory of of "vegetal" stabilizations carried out the past 15 years shows that several approaches are used : purely vegetal (21 %), casings (14 %), "tufts" (16 %), combs and fascines (16 %), planted cobblestones (23 %) and various combinations of these (10 %). The approach is positive and deserves to be analyzed, in order to initiate a shift towards those techniques that rely more strictly on plants, and to improve the techniques themselves (selection of species, design, adequation with local conditions, biodiversity, integration into the landscape, etc.). It is the reason why the monitoring of sites is essential. Until now, however, the information available on their operation is very incomplete and dispersed, neither pre- nor post-setup studies having been undertaken, for exemple. A couple of MSc theses in biology (Legrand, 1997 ; Lenoir, 1998), completed at the URBO, have gathered and summarized this information under the form of an epistolar census mailed to the various districts involved, and analyzed a number of sites. These studies point to the interest of using more structured approaches, that is analyze the situations of sites before the works are undertaken, and monitor them for a sufficiently extended period thereafter (ca. 5 years). On the basis of current projects, a number of sites representative of the various techniques have been place under surveillance in a both ecological and geomorphological perspective (ULg).