Blood stock availability at the Blood Donation Unit (UDD) PMI Langkat Regency is often disrupted due to unpredictable demand and reliance on voluntary donors, leading to risks of shortages or surpluses, especially for certain blood groups. This study applies a data mining method using the Single Moving Average (SMA) approach to predict the demand for blood bags based on blood groups (A, B, AB, and O) over the past year. The forecasting process calculates the average demand over a specific period and measures accuracy using MAPE, MSE, and MAD. The results show that the SMA method provides reasonably accurate predictions, with blood group O having the highest average demand of 161 bags per month, followed by blood groups A, B, and AB. An average MAPE value below 10% indicates that this method is effective for blood stock planning at UDD PMI Langkat, helping to optimize blood inventory management and minimize the risk of shortages or surpluses.