This paper analyzed the fundamental limitations of previous work and developed a new method to optimally locate actuators and sensors for structures with close modes. Optimization criteria were defined based on the distinguishing modal controllability and observability measures of close modes. An appropriate genetic algorism was adopted as the optimization algorism. Solving the high order Lyapunov functions was avoided by means of the closed-form expressions for controllability and observability Grammians. Since structure with widely separated natural frequencies is approximately balanced, computational efficiency was improved by grouping close modes together and dealing with the resulting subsystems independently. Finally, the effectiveness and optimality of the novel placement scheme were verified on a model structure with close modes. This paper analyzed the fundamental limitations of previous work and developed a new method to optimally locate actuators and sensors for structures with close modes. Optimization criteria were defined based on the distinguishing modal controllability and observability measures of close modes. An appropriate genetic algorism was adopted as the optimization algorism. Solving the high order Lyapunov functions was avoided by means of the closed-form expressions for controllability and observability Grammians. Since structure with widely separated natural frequencies is approximately balanced, computational efficiency was improved by grouping close modes together and dealing with the resulting subsystems independently. Finally, the effectiveness and optimality of the novel placement scheme were verified on a model structure with close modes.