The current research deals with short term forecasting of demand on Blood material, and its' problem represented by increasing of forecast' errors in The National Center for Blood Transfusion because using inappropriate method of forecasting by Centers' management, represented with Naive Model. The importance of research represented by the great affect for forecasts accuracy on operational performance for health care organizations, and necessity of providing blood material with desired quantity and in suitable time. The literatures deal with subject of short term forecasting of demand with using the time series models in order to getting of accuracy results, because depending these models on data of last demand, that is being stable in short term. The aim of research is decreasing forecasting' errors of demand on blood units & plasma for period (2005-2007) through using three seasonal quantitative models for forecasting, these are multiplicative, additional, and Winters models, and choosing model that performs lowest amount from measures of forecast' errors, these are mean squired error, mean absolute deviation, and mean absolute percentage error.
The results of the research showed achievement the multiplicative model which is used in forecasting of demand of blood units, and the additive model which is used for plasma, to lowest amount of forecast' error, and the recommendation was using these models for forecasting of demand.