By utilising the Italian component of the Image & Corine Land Cover 2000 project (I&CLC2000) it has been proposed the creation of a database containing vegetation spec -tral signature, by using Landsat 7 ETM+ images acquired during the summer of 1999-2001 and covering the whole Italian territory, in order to contribute on the improving of the land cover/use classification processes. The initial project structure foresees both the spectral and statistical results analysis of beechwood land cover (CLC2000 220.127.116.11 class). For the spectral analysis it has been utilised the vector layer retailed from the CLC2000 and ap -plied on the Landsat images. The statistical analysis on the selected areas has been effectu -ated by non-parametric test (Kruskal-Wallis ANOVA). In order to implement the spectral dataset of land vegetation component, this study aims at the chestnut analysis (CLC2000 18.104.22.168 class), by using the same methodology previously applied to the beachwood.