Accurate tidal datum transfer is crucial in hydrographic surveying and coastal engineering, particularly in regions characterized by complex tidal regimes. This study evaluates the performance of tidal datum transfer methods in a mixed semi-diurnal environment using simultaneous water level observations from a reference and a secondary station situated in the Persian Gulf and Strait of Hormuz. This specific area was selected due to its unique hydrodynamic conditions, shallow bathymetry, and significant non-linear tidal interactions that complicate conventional datum establishment. The classical Range-Ratio approach is systematically compared against Grant’s parametric least-squares method, which mathematically estimates a scale factor, a time shift, and a datum shift by fitting the continuous tidal curve. To assess the robustness, accuracy, and sensitivity of these methods to observation lengths, a moving-window analysis was implemented over record lengths of 0.5, 1, 3, 7, and 15 days using 10-minute shift steps. Initial harmonic analysis of the observed data yielded form factor values of 0.38 and 0.43, confirming a mixed, predominantly semi-diurnal regime that theoretically satisfies the prerequisites for applying the Range-Ratio method. However, evaluation based on the Root Mean Square Error (RMSE) across all tested windows demonstrated that the parametric least-squares method is significantly more stable and reliable. By utilizing the entire continuous record rather than discrete data points, Grant’s method mitigates the inherent limitations of the classical approach, which relies solely on a limited number of high and low water extrema and remains highly sensitive to short-term meteorological noise and shallow-water effects. Ultimately, the findings strongly advocate adopting continuous curve-fitting techniques combined with moving-window analyses for precise hydrographic datum transfers in complex coastal basins.