In two field experiments with cotton (Gossypium), the relation between moderate soil water depletion and leafair temperature difference (ΔT) was investigated for feasibility of measurement and possible use as a guide to irrigation scheduling. Hourly temperatures of the upper leaves and of the air 1 m above the canopy were obtained for three cultivars of short‐staple cotton (G. hirsutum L.) and one of long‐staple (G. barbadense L.). In addition, hourly measurements were made of the vapor pressure 1 m above the crop. These data were taken during six irrigation cycles extending over two summers and encompassing a wide range of saturation deficit (SD).In contrast to previous experiments in which a severe drought had been developed deliberately to intensity plant response, the experimental plots were irrigated on a schedule to obtain a high lint yield.The temperature differences were measurable and predictable. When the SD remained steady from day to day, the mean daytime ΔT decreased 1 C on the day of irrigation and an additional degree the next day owing to leaf rehydration and consequent stomatal opening. Conversely, this 2 C decrease upon irrigation reflects the rise in ΔT that had occurred during the previous drying cycle. When the SD decreased appreciably, the stomatal opening caused by irrigation was masked. Because the SD seldom remains steady over long periods, a measurement of ΔT alone is not a sufficient indicator for irrigation. Thus, allowance was made for such changes in SD by calculating a regression of ΔT on the SD from data taken from a recently irrigated crop.Although the use of ΔT to schedule irrigations seems feasible, the method is restricted primarily to regions with sunny climates. Even here it requires accurate, wellreplicated leaf temperature measurements as well as measurements of air temperature and vapor pressure.