This study was performed to test three methods based on the FAO-56 ‘‘dual'' crop coefficient approach to estimate actual evapotranspiration (AET) for winter wheat under different irrigation treatments in the semi-arid region of Tensift Al Haouz, Marrakech (center of Morocco). The three methods differ in the calculation of the basal crop coefficient (Kcb) and the fraction of soil surface covered by vegetation ( fc). The first approach strictly follows the FAO-56 procedure, with Kcb given in the FAO-56 tables and fc calculated from Kcb (No- Calibration method). The second method uses local Kcb and fc values estimated from field measurements (Local-Calibration method) and the last approach uses a remotely-sensed vegetation index to estimate Kcb and fc (NDVI-Calibration method). The analysis was performed on three fields using actual (AET) measured by Eddy Correlation systems. It was shown that the Local-Calibration approach gave best results. Accurate estimates of Kcb and fc were necessary for FAO-56 ‘‘dual'' crop coefficient application. The locally derived Kcb for winter wheat taken at initial, mid-season, and maturity crop growth were 0.15, 0.90 and 0.23, respectively. The Kcb value at the mid-season stage was found to be considerably less than that suggested by the FAO-56. Similarity between the seasonal pattern of normalized difference vegetation index (NDVI) and Kcb showed potential for modelling NDVI into a Kcb. The obtained relationships between Kcb and NDVI, and between fc and NDVI could be easily incorporated within the FAO-56 ‘‘dual'' crop coefficient model and, thereby, provide a means to apply remotely sensed observation for real-time wheat irrigation scheduling. The results obtained were very acceptable especially when the soil evaporation is negligible. Therefore, the Kcb–NDVI relationship employed in the FAO-56 ‘‘dual'' crop coefficient model holds great potential for estimating crop water requirements on an operational basis and consumption at a regional scale.