Currently, about 90 percent of the population of developing countries live in arid and semi-arid areas. Therefore, it is necessary and inevitable to identify the key and effective factors on the growth of agricultural products . The purpose of this research is to evaluate the accuracy and efficiency of the latest version of the AquaCrop model for simulating seed yield, biomass and water productivity efficiency of safflower in the form of a randomized complete block design with one control treatment (Conventional irrigation) and four deficit -irrigation treatments, including (RDI80, RDI60: reduced irrigation adjusted by supplying 80 and 60% of soil moisture deficiency) and (PRD80, PRD60: fixed partial root-zone drying irrigation with supply of 80 and 60% of soil moisture deficiency) and with four repetitions in Khuzestan University of Agriculture and Natural Resources was conducted. In order to evaluate the reliability of the model, some statistical indicator including the average normalized root mean Square error (NRMSE), the Coefficient of Residual Moss (CRM), Coefficient of Determination (CD), maximum error percentage (ME), Wilmot index (d) and the modeling efficiency (EF) was used. The statistical index (NRMSE) was obtained for seed yield (4.14), biological yield (10.16) and water productivity efficiency (4.9) and showed that the accuracy of the model in simulation is excellent. The results of the model validation show the accuracy and efficiency of the model in simulating seed yield, biomass and water productivity efficiency, and therefore, this model can be used to manage safflower irrigation in different states of deficit irrigation.