Winter oilseed rape (Brassica napus L.) has a poor nitrogen use efficiency (NUE). Among several approaches to increase NUE site-specific N-application is one promising attempt. Crop parameters like green area index (GAI), dry matter (DM) or total nitrogen amount of the shoot (total N) may give useful informations about the local variation of N supply and therefore can be used to find optimal local N rates. Crop parameters can be estimated by measurements of canopy reflectance leading to vegetation indices (VI), which combine reflection values within discrete wavebands in different mathematical forms. There are already VIs suggested for winter oilseed rape, which are suited to estimate the plant parameters mentioned above to only a limited extent. In course of this work, more diverse VIs were generated by three different methods and were systematically screened for their suitability to estimate the plant parameters. Therefore, wavebands in a spectrum of 400 nm to 1000 nm in 10 nm steps were used. First, the simple ratio form (SR) λ1/λ2, second, the normalized difference index form (NDI) (λ1-λ2)/(λ1+λ2) and third, a multiple regression (MR) was used to find the best linear combination within available bands. These methods were tested on data obtained by a 3-year field experiment, 2004-2006, which was conducted on an experimental farm near Kiel, Germany. To validate the newly derived VIs an independent data set for the same location was used. Calibration and validation show, that VIs can be found, which better estimate the plant parameters mentioned above than the commonly used ones.