Automated irrigation scheduling of grain crops using a combination of plant and soil water sensing feedback has not been widely investigated. A three-year study was conducted at Bushland, Texas to investigate irrigation management of grain sorghum (Sorghum bicolor, L.), in 2012 using plant feedback with a single thermal stress threshold, and in 2018 and 2019 using multiple thermal stress thresholds and a combination of plant and soil water sensing (Hybrid) feedback. The goals of the studies were to optimize grain yield, crop water productivity (CWP) and irrigation water productivity (IWP) using sensor feedback at irrigation levels similar to 80 %, 50 % and 30 % (designated I-80, I-50 and I-30) replenishment of soil water depletion to field capacity as determined with weekly neutron probe readings (the "manual" method). Results in 2012 indicated that irrigation scheduling using plant feedback alone with a single thermal stress threshold produced grain yields that were significantly less (0.49 and 0.38 kg m(-2)) compared with the manual method (0.63 and 0.51 kg m(-2)) at the I-80 and I-50 treatment levels, respectively. However, in 2018, the Hybrid feedback method produced mean grain yields (0.87 kg m(-2)) that were significantly greater compared with the plant feedback (0.76 kg m(-2)) and manual (0.74 kg m(-2)) irrigation scheduling methods at the I-80 treatment level. In 2019, mean grain yields (0.86, 0.83 and 0.88 kg m(-2)), CWP (1.25, 1.29 and 1.20 kg m(-3)) and IWP (2.11, 2.19 and 1.88 kg m(-3)) for the Hybrid, plant feedback and manual methods, respectively, were similar at the I-80 level. These results suggest that plant and soil water sensing feedback using multiple thermal stress thresholds and watering levels have the potential to produce optimal crop response for grain sorghum. More research is required to test the efficacy of soil water sensing in combination with plant sensing for other crops.