In this paper we describe the use of genetic programming for the prediction of blood demands. As blood bags for Hospitals are provided by blood banks on demand, predicting the needed amount of those should be as precise as possible. In order to achieve such an accurate prediction we have used genetic programming for data based modeling in order to find a mathematical model which predicts the blood bag demand of a hospital. This model should allow the hospital to minimize storage costs and the probability of running out of certain types of blood bags. In addition to the anonymized patient data provided by the General Hospital Linz, Austria we have also considered supplemental data such as weather and historical data such as the blood demand of the last few days which might lead to a more accurate model.