Performing forecasting of needs in the subject of research for 2021. The subject of the research is an institution operating in the territory of the Republic of Poland. The subject of the research are planned needs and expenses incurred in the food department in PLN. By comparing the planned needs and the expenses incurred to cover them, their imbalance was observed. This became the basis for further research in terms of forecasting expenses in order to identify future needs. The selection of forecasting methods was preceded by the performance of multidimensional comparative analyzes and their evaluation. The research began with the search for regularities in the form of trends and seasonality by analyzing and assessing the time series of food department expenses. The analyzes revealed the lack of clear regularities in the form of the trend and seasonality governing the time series of expenses in the food department from January 2014 to November 2020. There was only a slight downward trend and weak / not clearly visible seasonality on a monthly basis. Practical implications: The detection of regularities made it possible to select methods for forecasting expenses in the food department for the future, which include: Holt ‒ Winters and the naive method. A comparative analysis of two forecasting methods allows for the selection of the best one. The practical implication is that future needs can be calculated by forecasting incurred expenses. Forecasting models available in the literature can / should be used to determine future forecasts. Planning needs by forecasting expenses incurred to secure them in dynamic terms.