Given the global importance of sugarcane, its monitoring is strategic. There is a huge amount of agrometeorological and environmental data generated by satellites that can be used for this purpose bringing countless advantages, such as economy, agility and precision in the analysis. This paper analyses the utility of the time series of remote sensing data in the generation of numerical models for estimate the sugarcane production through multiple linear regression analysis using the variables NDVI / MODIS, WRSI and planted area. The study area was formed by seven municipalities in the state of Sao Paulo, Brazil, the first national sugarcane producers. The good models accuracy was shown by the correlation coefficient (R2) above 0.94 for all generated models. We have identified that the time series of remote sensing data were useful for longer periods of data rather than each individual crop season considered.