In a tropical country such as Indonesia, dengue fever has become an endemic outbreak because it occurs almost every year, especially in the beginning and towards the end of the rainy season. Districts affected by dengue fever need accuracy in predicting the number of patients and drugs. Hospitals need a suitable drug order inventory system to treat dengue patients. The objective of this paper is to develop a model with higher accuracy and suitability. We propose a two-phase model with several steps and methods to fulfill the objective of each phase. In first phase, we determine a drug forecasting method to predict the number of patients and drug demand. In the second phase, we determine the drug order system (continuous review and periodic review) including the optimal quantities, frequencies, safety stock, and reorder point. The proposed model was then applied in a hospital in Sidoarjo. The results show that the proposed model can be applied in the district and the hospital; and the drug demand forecasting and ordering system resulted in optimal inventory and costs. This study proposed a new multi-phased model that can be used in drug demand forecasting and ordering system, with practical benefits for hospitals and the healthcare systems.