With the rapid growth of the global population and the increasing demand for food production, unprecedented pressure has been exerted on limited water and soil resources. This challenge is particularly critical in arid and semi-arid regions such as Iran, where efficient and intelligent water resources management has become a vital necessity. The main objective of this study was to develop and evaluate mathematical models for estimating the cumulative water requirement of wheat as a function of time within homogeneous agro-ecological zones (AEZs) at the national scale, in order to provide a simple yet accurate tool for irrigation planning. To this end, water requirements were calculated using data from the National Crop and Orchard Water Requirement System based on the FAO Penman–Monteith standard equation. Given the sigmoidal (S-shaped) pattern of cumulative water requirement changes over the growing period, three mathematical models—namely the three-parameter logistic, Gaussian, and third-order polynomial models—were selected and tested across 12 different agricultural classes in the country. Model performance was evaluated using the coefficient of determination (R²), normalized root mean square error (NRMSE), and Nash–Sutcliffe efficiency (NSE). The results demonstrated that all three models exhibited very high accuracy across all agricultural classes. However, further technical analysis revealed that despite their comparable statistical accuracy, the logistic model is more practical and suitable than the Gaussian and third-order polynomial models for continuous modeling and estimation of cumulative wheat water requirements, as it preserves a physically and biologically meaningful form throughout the entire growing period. These models enable accurate estimation of water requirements at any time interval during the growing season and represent an effective step toward achieving sustainable agriculture and improving water use efficiency across different regions of the country.