Infrared thermometry was applied to estimate the canopy temperature of apple trees with the aim to detect a water stress condition early by remote sensing. The measurements were taken in Michigan during Summer 1998 in a 4-year-old apple orchard. Digital thermo-images of the canopy were taken using a IR imaging radiometer on well-watered trees and trees in a water shortage condition. The images were taken considering the geometrical relationship among camera position, canopy, and sun position. During the measurements, environmental (air and soil) conditions were also monitored. A software program was developed to analyze the thermal data, to show the thermal frequency distribution and to estimate the statistical parameters, which are able to represent the physiological condition of the trees. An increase of the canopy surface temperature (connected to the partial stomatal closure that is affecting the leaf energy balance) was detected early in the non-irrigated plants, compared to the well-irrigated trees, already when physiological responses as photosynthetic activity and fruit growth were not yet negatively affected by water deficit. The study confirms that there are the theoretical basis to use infrared thermometry and digital image processing to early detect the water stress on fruit trees.