In this study, the capability of crop water stress index (CWSI) based on satellite thermal infrared data for estimating water stress and irrigation scheduling in sugarcane fields was evaluated. For this purpose, eight Landsat 8 satellite images were acquired during the sugarcane growing season (May–September 2015). Simultaneous with the satellite overpass times, in-situ measurements of canopy temperature and vegetation water content (VWC) were conducted in forty points located in eight sugarcane fields per image (in total 320 observation points in 32 fields). These fields were selected with different ages (Plant, Ratoon 1, Ratoon 2, and Ratoon 3) and irrigation schedule. Then, the CWSI was calculated in three different ways including: 1) based on the Idso method and using the handheld infrared thermometer, 2) based on the Idso method and thermal infrared data of Landsat 8 satellite imagery, 3) using a new proposed procedure for retrieving CWSI from the satellite imagery with using the hot and cold pixels. Results show a good relationship between the calculated CWSI based on field measurement and new CWSI based on satellite data with the coefficient of determination of 0.49–0.85 and the root mean square error (RMSE) of 0.12–0.29 for different images. Further, a negative relationship between VWC and CWSI, with R2 values of 0.42–0.78, was observed. This relationship increases with developing sugarcane canopy, and decreases with an increasing plant age. Comparing recorded irrigation events in the fields, estimated CWSI and VWC shows that water stress can be classified into three critical classes including high water stress (0.70 < VWC ≤ 0.75), medium water stress (0.75 < VWC ≤ 0.8), and low stress (0.8 < VWC ≤ 0.85). This classification can be used as a part of an operational procedure for appropriate irrigation scheduling. All of the aforementioned results indicate that the CWSI based on the proposed approach in this study can be used effectively for monitoring water stress and irrigation scheduling in sugarcane fields using satellite imagery without any need for ground ancillary data.