In an ongoing multi-year study, the potential for using aerial hyperspectral (HS) images combined with aerial thermal images to evaluate and map water and nitrogen (N) status in potato fields was investigated. The paper describes procedures carried out on potato plants to evaluate N concentration based on spectral measurements and aerial HS images in the VISNIR range and water status based on ground thermal images. Field experiments were conducted to create wide ranges in N concentration and water status. The usefulness of partial least squares regression (PLSR) for spectral analysis to predict N concentration at the leaf scale was described. PLSR analysis obtained 1:1 prediction models of leaf-%N and petiole NO3-N (R2 of 0.86 and 0.74 respectively). The multi-dimensional beamlet-decorated recursive-dyadic-partitioning (BD-RDP) was introduced and applied to a 210-band aerial HS image of the experimental plot. The multi-scale segmentation significantly reduced noise in the original image and successfully uncovered spatial structures in the image according to the N treatments. The use of ground thermal images for water status estimation in potato plants was shown to be feasible: Temperature was negatively and strongly correlated with stomatal conductance. To the authors' knowledge, this is the first attempt that HS and thermal images have been used simultaneously for N and water status estimation for potato.