Abstract: Characteristics of given data mining tasks have not been considered in present universal integration model of OLAP and data mining, and this reduces their efficiency. This paper proposes a integration model of OLAP and mining association rules by analyzing the characteristics of mining association rules and OLAP, and gives its algebraic structure description. Based on this work, a theory of mining association rules based on OLAP is developed. It simplifies mining steps and enhances its efficiency. The proposed model and its theory are approved to be correct and feasible by experiment using some supermarket real data.