The balance-sheet method has been widely used for decades as a reference method to calculate the amount of nitrogen (N) fertilizer to supply to winter wheat. With this method, the amount of N fertilizer is calculated at the end of the winter to support the N requirement for an estimated yield level. This method is now being questioned because it has become difficult to predict yields, given climate change and the increased presence of pests in agroecological cropping systems, resulting in greater uncertainty about the yield achieved compared to conventional systems of the past. Alternative methods, such as the innovative approach of nitrogen fertilization (APPI-N; Ravier et al., 2018), are proposed for dynamic fertilization management: the amount of nitrogen fertilizer is adjusted to the real nitrogen status of the plant using the N nutrition index (NNI). Lemaire et al. (1989) define the NNI as the ratio of the measured nitrogen concentration (%N) to the critical nitrogen concentration (%Nc; Justes et al., 1994); it can be indirectly determined by measuring the leaf chlorophyll content using hand-held nondestructive contact devices (i.e. N-Tester, SPAD). However, their use is time-consuming and not suitable for routine use by farmers or advisors. Different optical measurements by proxi- or remote sensing using leaf color charts or vegetation indices combining visible and infrared wavebands at leaf or canopy level seem promising for estimating the nitrogen status of winter wheat. In the literature, a dark green color index (DGCI), was specifically created for the purposes of leaf nitrogen concentration analysis using visible images (Fig.1). The range of DGCI values is between 0 (very yellow; low N rate) and 1 (dark green; high N rate). Indeed, nitrogen is directly related to leaf color because it is a key component of the chlorophyll molecule (Tracy et al., 1992). First studied on grass (Karcher and Richardson, 2003), it was then successfully tested on maize crops (Rorie et al., 2011), provided that the images were normalized to correct for differences in lighting conditions during and between experiments and differences between cameras. In order to better understand the relationships between DGCI measurement methodology and nitrogen fertilization, different experiments were conducted at a canopy level. The objectives of this work were to explore the potential of the DGCI in winter wheat crop at canopy level for estimating NNI deduced from conventional methods (N-Tester). The trial zone represented 24 microplots: half of them were seeded with a winter wheat (Triticum aestivum L.) cv. LG Absalon and the other half with a mixture of cultivars (LG Absalon, Lipari, Rubisko, Tenor). To create a nitrogen gradient, the plots were fertilized according to five modalities (with two replicates) corresponding to different nitrogen doses ranging from 0 to 300 kgN.ha-1, with a nitrogen index ranging from 0.4 (minimal N status) to 1.2 (supra optimal N nutrition). A series of non-destructive measurements was performed on this N gradient during two growth stages, 2 nodes and heading. A chlorophyll-meter (N-tester) and three different optical systems (a digital camera and two smartphones) were used to estimate indirectly the leaf chlorophyll concentration from NNI.To validate protocol for image acquisition, three factors were tested: 1) DGCI vs. relative DGCI to evaluate the impact of standardization of image data 2) different optical systems to evaluate their impact and that of the white balance (manual or automatic) on the quality of the results 3) a pure wheat cultivar vs. mixture of cultivar to evaluate the capacity of the DGCI face to different cultivars. Statistical analyses were conducted to investigate the correlations between the N-Tester measurements and those deduced from images (DGCI and relative DGCI) using R software and RStudio, an integrated development environment for R. First results demonstrated that the relative DGCI values were highly correlated with NNI measurements with a significant degree of association. Best results were observed considering only one cultivar (LG Absalon). The “manual” option of white balance improved the results. Furthermore, there seemed to be a varietal effect depending on whether a mixture or a single cultivar is studied. As the results were only obtained in one year and at one location, further measurements are needed to validate this new methodology, which is easy to use and can be carried by any type of mobile platform (smartphone, drone, robot, etc.). From the results, the DGCI method developed on a large scale could be used as a nitrogen nutrition indicator on arable crops for dynamic monitoring of nitrogen fertilization.