The agroecological zone model proposed by FAO (ZAE/FAO) is an important tool in agricultural planning and can be used in Jataí, which has been gaining space in the production of sugarcane. The objective of this study was to compare the observed meteorological databases (BDMEP) with the estimated ones (Xavier and NASA POWER), in the estimation of potential productivity (PP) and attainable (PA) of sugarcane and its penalties for management (PMJ) and water deficit (PDF), using the ZAE/FAO model in the municipality of Jataí - GO. To this end, the BDMEP, Xavier and NASA POWER database, 33 years old (1984-2017), were used. Using BDMEP, Xavier and NASA POWER as input in the ZAE model, we observed an average PP of 259±11, 275±10 and 267±11 Mg ha-1 as, respectively. PA reaches 200 Mg ha-1 when using BDMEP in ZAE, with 50% of the crop year with PA above 180 Mg ha-1. There was an average penalty of 114.01±6.9 for management and 85.67±4.08 Mg ha-1 in stem yield. The meteorological variables of Xavier's database can be used in the ZAE model in the prediction of productivity penalty for water deficit and sugarcane management, with 63% and 88% adjustment to BDMEP, respectively. NASA POWER can be used in the simulation of penalty for failures in sugarcane management, with up to 87% approximation of those observed. Both sources when used in the prediction model assist in agricultural planning.