Demand forecasting studies are one of the strategic issues which are considered as significant for academicians and decision makers. Constantly changing political, cultural, legal and economic developments have pushed the companies to predict under uncertainty. It is very difficult to model some problems with traditional methods in complex, multi-dimensional and highly uncertain environment. Fuzzy approaches provide an easier modelling owing to their flexible nature. In this study, a demand forecasting study in a FMCG (fast-moving consumer goods) company has been applied using a rule based fuzzy logic approach which utilizes fuzzy set theory, fuzzy if-else rules and fuzzy inference concepts. The obtained results have been compared with real demands and low MAPE (mean absolute percentage error) values have been calculated. Also the obtained results have been compared with time series approach and the superiorities of the methods are discussed.