Use of new tools to prescribe Nitrogen (N) fertilization rates based on wheat canopy reflectance is promissory in areas where rainfall is certain during late tillering and jointing. The use of canopy spectral reflectance indexes (NDVI) using a GreenSeeker sensor has been successfully used to prescribe N fertilization in several wheat areas around the world. The objective of this study was to adjust and to validate a regional sensor based wheat yield predictive model for Argentina, and to evaluate a sensor based nitrogen rate calculator (SBNRC).  We evaluated two growing seasons (2005 and 2006) at INTA Paraná Research Station (Argentina, 31º 50S; 60º 31W). Three experiments (RCBD) with irrigation, 6 cultivars, and`several N rates (0 to 200 kg N ha^-1) as source of variation, were used to develop a sensor based yield predictive model. Four strip trials with N rates at planting were used to evaluate a local adjustment of the SBNRC.  Urea was used as N source in all cases. NDVI readings were obtained at first visible node (GS 30-31). A unique yield predictive model, was fitted polling all sources of variation, obtaining a satisfactory prediction (R² = 0.78). Average grain N concentration was 1.953%. Nitrogen availability and site variability affected strip sensor readings, N rate prescriptions and, N use efficiencies (NUE). The SBNRC did not increase grain yields, however NUE was significantly improved. Further experiences across a broader range of environments are still needed to validate the local adjusted algorithm.