Abstract: To improve the search efficiency and classification accuracy of traditional associative classification methods, an associative classification approach is proposed based on the clonal selection algorithm. In this algorithm, the rule-search and rule-selection processes are integrated to optimize and select the association rules in the evolution of the immune cell population while the associative classifiers are optimized as a whole. As a stochastic optimization algorithm, the proposed approach can deal with huge search space of association rules without necessarily generating all possible rules. The results of the simulation experiment show that the proposed algorithm achieves a good runtime and accuracy performance compared with conventional associative classification algorithms.