Calibration, validation and sensitivity analysis of the AquaCrop model for sugarcane varieties Plant growth simulation models are being increasingly used to assess the effects of water stress on sugarcane production (yield gap) and to optimize water use under limiting conditions, to improve the sustainability and profitability of rainfed and irrigated production. Thus, the objectives of this study were: i) to calibrate and validate the AquaCrop model and its input variables for direct estimation of the productivity of sugarcane varieties and evaluate its performance in comparison with experimental observed data; ii) to perform sensitivity analysis with the variables : plant density, canopy growth coefficient (CGC), canopy decline coefficient (CDC) and canopy full expansion coefficient (KcTrx) in order to identify the variables most impact on the simulation of the crop yield; iii) perform crop yield simulations using the model in order to apply it to different Brazilian locations and; iv) estimate the required irrigation blade in each location. The AquaCrop model was efficient in the calibration of the different sugarcane cultivars in the treatment with 100% water replacement, with high rates of accuracy, agreement and modeling efficiency (d, C ≥ 0.85 and E ≥ 0.7), which provided average absolute error below 10 tons ha-1. In the validation process AquaCrop model adequately simulated (d, C and E ≥ 0.7) the variables stalk productivity, raw sugar yield and dry biomass production of the 50 and 75% water replacement depths after calibration of the varieties on the 100% depth. AquaCrop model did not adequately estimate the water productivity of the sugarcane varieties studied obtaining low statistical indexes, such as accuracy index below 0.7 and negative modeling efficiency, the concordance index showed values close to zero. The most sensitive parameter to the variations tested was the crop coefficient with complete canopy expansion (KcTrx) followed by the canopy growth coefficient (CGC), right after the crop decline coefficient (CDC) and finally, the plant density. The reduction of the canopy growth coefficient (KcTrx) provided up to 50% drop in stalk productivity, raw sugar yield and crop perspiration, while the positive variation provided up to 5% increase in these variables. In the analysis of the density parameter of plants the decrease in the value of the parameter caused a reduction of about 3% in the yield of stalks and raw sugar yield and about 2% in the perspiration of the crop. The AquaCrop model showed good performance in estimating sugarcane stalks productivity in different regions of Brazil. There was an slight difference of the attainable productivity in at least half of the regions studied compared to the real productivity, so the introduction of a correction factor was proposed for a better estimative of stalks real productivity. The northeastern regions of Petrolina-PE and Aracaju-SE were the most responsive to irrigation technology (> 100%), requiring irrigation depths of 1700 and 800 mm respectivel...