This research addresses the problem of structural damage detection using linear vibration information contained in frequency response functions. The structural damage identification method is stated as an unconstrained optimization problem, which minimizes the error between measured and analytically computed vibration signatures using minimal measurement information. Two types of genetic algorithms are implemented to solve the structured and unstructured optimization problem of damage detection. The proposed algorithms are evaluated on case study simulations for different types of structures with increasing complexity. Noisy measurements are included in the simulations to investigate their effect on damage detection accuracy. The proposed method is compared to existing damage detection algorithms using accuracy measures that are based on Euclidean geometry. Case study results show that the proposed damage identification method is robust even in noisy measurement environments. A methodology for optimizing excitation and sensor layouts used for detecting damage in structures is presented by applying multi-objective genetic algorithms. In sensor layout optimization, the objectives are to reduce the number of sensors required while trying to increase the amount of information contained in the vibration signatures. Several case studies were investigated to determine the effect of using the optimum sensor layout designs evolved on the performance of the FRF-based damage detection method. The results show that the quality of the measurement information increased when the optimal sensor locations were used and the ability of the damage detection method to uniquely identify damaged elements was enhanced.