Actual evapotranspiration (ETa) is the important component of the water balance in arid and semi-arid basins, and its quantitative estimation is essential for sustainable water resources management. In this study, the spatiotemporal distribution of ETa across the Hirmand (Helmand) River Basin (~280,000 km²) was estimated for 2023 using the Surface Energy Balance Algorithm for Land (SEBAL) implemented within the Google Earth Engine (GEE) cloud computing platform. Input datasets comprised MODIS surface reflectance (MOD09GA), land surface temperature (MOD11A1), and ERA5-Land atmospheric reanalysis products. Mean annual ETa across the basin was estimated at 670 mm/year, following a clear decreasing gradient from the mountainous headwaters (>1,100 mm/year) to the arid lowland plains (<100 mm/year). Seasonal ETa peaked in spring (745 mm/season) and approached zero in winter. Correlation analysis revealed that net radiation (R = +0.703) and land surface temperature (R = −0.933) exhibited the strongest positive and negative associations with ETa, respectively, while NDVI (R = +0.372) and surface albedo (R = −0.053) showed comparatively weaker relationships. Validation against the MOD16A2GF product yielded a correlation of r = 0.858 and a coefficient of determination of R² = 0.736, confirming the reliability of the spatial distribution pattern. The systematic overestimation of SEBAL relative to MOD16 (PBIAS = +283.9%) is attributed to differences in the computational approaches of the two models and to the well-documented tendency of MOD16 to underestimate ETa in arid environments. The findings of study provide a robust quantitative framework for water resources management and transboundary water diplomacy in the Hirmand Basin.