The manual preparation strategy of a promosi package can encounter difficulties in determining the right product to be promosited caused of no sales data analysis and the large size of sales database.The solution to this problem is to apply data mining science to find buying patterns that consumers often make in a collection of sales transactions, so that company can create the right promosi packages to encourage increased sales turnover. The study scanned the sales database obtained from the UCI maching learning repository and intelligent system dataset with 12,224 sales transaction records and 126,898 record of goods sold. Furthermore, the data mining process carries out the process of constructing an FP-Tree tree structure, constructing a conditional FP-Tree and TID List, then extracting a combination of items and calculating support (S), confidence (C) and sorting association rules based on the value of S x C descending, from the highest to the lowest value. The combination of FP-Tree and TID-List algorithms can be used to help formulate promosi package strategies, by analyzing the sales database and finding the most frequently sold combinations of items.