The increasing population and water demand in North Africa, particularly in semi-arid and arid regions like Morocco, necessitate efficient water resource management. Morocco experiences significant water scarcity due to persistent droughts and rising agricultural water demand. Accurate estimation of crop water requirements, particularly crop evapotranspiration, influenced by climatic and crop-specific conditions, is vital for improving water use efficiency. This study focuses on wheat, a staple crop in Morocco's Draa-tafilalet region, encompassing Ouarzazate, Zagora, and Tinghir. Traditional methods for estimating the crop coefficient require comprehensive climatic datasets, often unavailable. Therefore, this study aims to develop and validate artificial neural network models to predict monthly \(Kc \) values for wheat using limited climatic data. The results showed that the optimal model for predicting crop coefficients used a minimum and maximum temperatures, along with solar radiation, applying different configurations of hidden neuron layers for each region: (7, 5) for Ouarzazate, (8, 6) for Zagora, and (9, 6) for Tinghir. The best model's accuracy was close to 1, showing a statistically significant accordance between observed and predicted values. This study highlights the model's high accuracy and recommends its use for predicting crop coefficients values with limited climatic factors. It also supports water users in developing and updating \(Kc\) databases for each region based on climatic conditions.