Measuring plant bio-physiological parameters in a field can be expensive and time consuming. Vegetation indices from proximal and remote sensing can be used to estimate variables as well as to overcome the problem of entering in rice paddies. This study aims to explore the potential of vegetation indices (i) to identify presence of of weed and plant stand (ii) to assess Leaf Area Index (LAI) in rice paddies in Italy. The study field was located in south Sardinia (Italy).  The experiment was carried out on three different rice cultivars during 2010 and 2011 summer. A time series of FieldSpec canopy reflectance data and multispectral satellite remote sensing imagery were acquired during both growing seasons.  Analysis were carried out to compare the performance of thirty vegetation indices. In 2010 growing season, the CRM and MTCI (derived from new red-edge spectral band) proved to be the best predictors of green LAI (R²=0.72), while the traditional CGM (R²=0.67) index showed good linearity with green LAI.  In 2011 growing season, the heterogeneity of seed density and the strong presence of weeds (despite the weeding control) caused the lack of correlation between LAI and vegetation indices since the beginning of the growing season but additional insights of the location of weeds was obtained.