The genetic variability of apple tree response to drought is scarcely known because characterization of response to water constraint is usually low-throughput, which limits the possibility to take this trait into account in breeding programmes. In this research phenotypic response of apple tree scion hybrids of a 'Starkrimson' × 'Granny Smith' adult population to soil drought was considered through the leaf temperature as a proxy for transpiration, each tree being grafted onto the same rootstock genotype. The experiment was located in an experimental orchard where trees were submitted to contrasting summer irrigation treatments. The use of airborne multispectral images (in reflective and thermal infrared spectral domains) was explored as a possible approach to evaluate the hybrids' behavior in orchard conditions. Emphasis was put on the leaf transpiration status which was approached by measuring foliage temperature in thermal infrared (TIR) and by calculation of the Water Deficit Index (WDI). Here we proposed to calculate the WDI at tree scale, to overcome limitations of TIR interpretation for non-covering vegetation by taking the crop cover fraction into account. This was performed through the calculation of the Normalized Difference Vegetation Index (NDVI). Our results highlight capacity of thermal imaging for phenotyping water stress. Spatial resolution of TIR images needs to be improved in order to refine the phenotyping data and specific image treatment is required to extract relevant leaf thermal information which can allow for effective differentiation of hybrids' behavior in response to soil water constraint. Remote sensing thus appears promising for phenotyping varietal response of apple tree to soil drought at adult stage in field conditions.