Abstract: Many real world problems involve the simultaneous optimization of various and often conflicting objectives. These optimization problems are known as multi-objective optimization problems. Evolutionary multi-objective optimization, whose main task is to deal with multi-objective optimization problems by evolutionary computation techniques, has become a hot topic in evolutionary computation community. The solution diversity of multi-objective optimization problems mainly focuses on two aspects, breadth and uniformity. After analyzing the traditional methods which were used to maintain the diversity of individual in multi-objective evolutionary algorithms, a novel nondominated individual selection strategy based on adaptive partition is proposed. The new strategy partitions the current trade-off front adaptively according to the individual's similarity. Then one representative individual will be selected in each partitioned regions for pruning nondominated individuals. For maintaining the diversity of the solutions, the adaptive partition selection strategy can be incorporated in multi-objective evolutionary algorithms without the need of any parameter setting, and can be applied in either the parameter or objective domain depending on the nature of the problem involved. In order to evaluate the validity of the new strategy, we apply it into two state-of-the-art multi-objective evolutionary algorithms. The experimental results based on thirteen benchmark problems show that the new strategy improves the performance obviously in terms of breadth and uniformity of nondominated solutions.