Abstract: Association rule mining, one of the most important and well-researched techniques of data mining, investigates the possibility of simultaneous occurrence of data. Given a collection of data, critical analysis can be made on the statistical frequency of data, relationship between different data items and their comparative frequency, etc. This study proposes a new algorithm Reverse Apriori Frequent Pattern Mining, which is a new approach for frequent pattern generation of association rule mining. The proposed algorithm works efficiently, when the number of items in the large frequent itemsets is close to the number of total attributes in the dataset, or if number of items in the large frequent itemsets is predetermined.