Mining of Association Rules between the items of a huge centralized Database is very interesting and important research area. Its importance becomes more significant in case of sales transaction. There are a number of algorithms working on this specialized research area. The algorithm presented in this study, (Centralized Mining of Association-Rules), CMA is more efficient than the previous existing algorithms, as it not only reduces the overhead of frequent disk I/Os and the CPU cost, but it also reduces the database scan to the half. The CMA algorithm, presented in this study, basically takes the best features of two state-of-the-art algorithms in the area, i.e., the technique of PARTITION algorithm of centralized database area is taken for partitioning the huge Database and then, the DMA algorithm of distributed database environment is applied on each partition. The large itemsets to be found at the end of operation at each partition are to be merged together and then the actual set of large items ets is finally created.