Production forecasting is one of the most critical stages in short-term planning in upstream oil companies. The oil volume that will be produced, called production target, directly influences all companies actions and impact critically their market image. Therefore, it is noticed the importance of using models to consider uncertainties to evaluate production forecasting risks. This stochastic approach has been done through a simulation model which consider four dimensions of variables: installed production potential, new wells entry, scheduled maintenance program, and operational efficiency. Among those dimensions, the new wells entry is one of the most sensitive to the simulation results, because of its high degree of uncertainty. Thus, this work aims to develop a fuzzy inference system to forecast the new wells production start date. The system is designed integrated to the simulation model in order to increase its accuracy. The results show that the fuzzy inference system can be used to forecast wells production start date and its use increases oil production forecasting accuracy.