AbstractCrop evapotranspiration (ETC) of rice under an alternate wetting and drying (AWD) irrigation practice may be different from conventional (CON) irrigation methods due to the drying (water stress) phase. The focus of this study was to identify a suitable approach for estimating ETC for a rice crop under AWD practice. Field lysimeters were used for estimating daily ETC under AWD and CON practices during monsoon and non‐monsoon seasons. Lysimeter‐based evapotranspiration (ETC_Lys) was compared with that resulting from the product of reference evapotranspiration (FAO‐56 Penman–Monteith equation), crop coefficient (FAO tabulated values‐KC_Tab and FAO equation‐KC_Eqn) and soil water stress coefficient (linear equation‐KS_Lin; FAO equation‐KS_FAO; logarithmic equation‐KS_Log). The KC_Eqn and KS_FAO methods were found to be better estimates for KC and KS, respectively, for AWD practice. The mean KC in initial, mid and late season for AWD were found to be very close to CON practice. The KS was observed to vary from 1 to 0.15 for AWD. Overall, AWD irrigation practice saved 10–20% irrigation water with about 2% reduction in grain yield and 13% reduction in ETC. The reduction in ETC indicated that KS must be considered when calculating the ETC in rice under AWD irrigation practice.