Blood transfusion is a process of sending or transferring blood to another place and the task is delegated to the PMI Blood Donation Unit with several tasks including; deployment and preservation of blood donors, provision and processing of blood, and distribution of blood to health agencies. However, the supply and demand from health agencies have a significant difference. The difference for each blood group is very large, in group O deficiency by 28%, in group A deficiency by 38%, in group B excess by 28%, and in group AB deficiency by 84%. To overcome this problem, it is necessary to estimate the demand for blood that will occur in the future period. One of the tools that can answer this problem is demand forecasting and what will be used in this study is forecasting using the Support Vector Machine (SVM) method. This SVM method is a relatively new machine learning-based technique for making predictions, both used in classification and regression cases. The result of this study obtained good MAPE values, namely in blood group O is 14%, in blood group A is 15%, in blood group B is 13%, and in blood type AB is 24%.