The irrigated area expansion in South Asia brings challenges to water resources management, and the combination of irrigation and climatic variability has caused periodic water shortages in many parts of the region. In this context, monitoring and predicting the influences of irrigation and water extraction on the water balance has become increasingly important for local governments and decision-makers. In this study, we apply a demand-driven irrigation scheme to an existing hydrological monitoring system that uses the Noah-MultiParameterization land surface model within the NASA Land Information System framework. Including prognostic simulation of irrigation generally improves the simulated ET compared to diagnostic satellite-based products, though uncertainty in observations limits the confidence of this conclusion. Results for soil moisture were mixed, with improved agreement with observation in Pakistan but degraded in parts of India. Including irrigation has a significant benefit in the simulation of groundwater depletion compared to in-situ and satellite-derived observations. Additionally, a sensitivity analysis is carried out to study the relative contribution of different input combinations to the hydrological cycle. These input combinations include choices of precipitation, soil texture, irrigation fraction map, groundwater fraction of irrigation, and greenness vegetation fraction (GVF) estimates. The results indicate that irrigation estimation and ET is most sensitive to irrigation fraction and soil texture datasets while the simulated soil moisture is more sensitive to soil texture datasets and the GVF approach. The study highlights the need to generate more consistent and reliable irrigation parameter datasets by incorporating local or remote sensing-based measurements for South Asia.