Abstract: Association rule mining is a way to find interesting associations among different large sets of data item. Apriori is the best known algorithm to mine the association rules. In this dissertation, clustering technique is used to improve the computational time of mining association rules in databases using Access data. Clusters are used to improve the performance of computer. Clusters are responsible for finding the frequent k item sets; hence lot of work is performed in parallel, thus decreasing the Computation time. This parallel nature of clusters is exploited to decrease the computation time in mining of data and also it reduces the bottleneck in the central site. Since after mining of data, there will be explosion of number of results and determining most frequent item sets will be difficult, so item sets are divided into two groups’ namely-globally frequent item sets and locally frequent item sets.