A spatial model is presented using logistic regression, fuzzy classification and GIS, to map the risk to the ravine. This model uses the maps: vegetation index, slope, and structural lineaments and roads density maps. The chance in favour of a gully event for these maps was estimated using odds ratio. The model calculates the risk of gully in each pixel, from the weighted sum of fuzzy maps, and uses r2 value obtained by frequency ravine points observed in each class of thematic maps as weight values.