While a lot of work has been done on the prediction of waterlogged soils in small catchments, we need to improve knowledge of the extension of these soils in large catchments. The spatial extent of water saturated soils is a key variable for the hydrological response of a catchment and more generally for hydrological processes (flooding, solute transport, erosion). Soil water saturation generates anoxic conditions recognized in soil morphology by redoxic or reductic features (WRB, 1998). Soil colour mottling, presence of iron and manganese (hydr)oxides, accumulation of organic soil material are morphological indicators which allow to infer the soil water regime. The delineation within landscapes of the area affected by topsoil redoximorphic features, called here Hydromorphic Zone (HZ), is therefore a major issue for soil survey studies (Soil Survey Staff, 1995). Due to close relation between topography and soil waterlogging, spatial modelling approaches based on topographic attributes have been largely developed (McBratney et al., 2003) to predict HZ distribution over space. At catchment scale, energy and material flow vary from sources to the outlet zone. The origin of deposits (alluvial or colluvial material) may change while progressing in the catchment. The aim of this study was to analyze the extent of redoximorphic soils in the stream neighbourhood in relation to the stream order. The study took place in a 10 000 km² catchment (the Vilaine river, Armorican massif, western France) where valley bottomland soils associated to hydromorphic characteristics are common and cover up to 20 % of the basin area. To describe the catchment organisation, we used the stream order according to the Strahler classification (1952) to test the behaviour of the topographic index (Beven and Kirkby, 1979) in a large range of morphological situations. Two methods were carried out to define the HZ extension in each valley bottomland according to stream order: (i) A field study based on the delimitation of the HZ according to redoximorphic features appearance along 60 transects; (ii) a modelling approach linking a DEM derived topographic index and the digitized stream network of the Vilaine. Considering topographic factors, the progressive valley expansion may represent an enhancing factor of HZ extension. Thus, simple topographic index modelling predicted increasing waterlogging in high order channel situations (order 6-7). By contrast, field delineation suggested that HZ extension remains stable with increasing order and decreases significantly for high order situations. Topographic index modelling appeared therefore effective in upper catchment situations (of first, 2nd and 3rd order). On opposite, the modelling efficiency was limited in high order situations where the indices proved inappropriate: in such context, interactions between adjacent hillslope and HZ are of second order importance whereas fine-scale valley bottomland morphology controls soil waterlogging duration. While progressing in the catchment, soil material nearby the streams shifts from colluvial origin in small orders to alluvial origin in high orders. The origin of soil materials affects soil water transmissivity and possibly plays also a major role in HZ extension. Finally, the integration of the stream order should considerably improve the efficiency of soil spatial distribution modelling over large areas.