Sales transaction data is a very valuable asset in business processes. Not only is it used to calculate profits and money, but large amounts of transaction data can also be used for various purposes to generate new knowledge (knowledge) in the transaction database. Ways that can be done for data processing and generate new knowledge from the data is to use data mining techniques. The technique used in this case is the FP-Growth Algorithm. The data structure used is a tree called FP-Tree. By using FP-Tree, FP-growth Algorithm can directly extract Itemset from FP-Tree. Research conducted by collecting data related to research in the case studio at Medan Haji Hospital Pharmacy where the variables taken are daily drug transaction data. The results of this study are part of the new knowledge of this sales data by applying the FP-Growth Algorithm that uses the concept of FP-Tree development in finding Frequent Itemset that is useful for the development of investment plans in the study areas taken. Keywords: Data Mining, Association Rules, Frequent Itemset, FP-Growth.