This paper presents an adaptive harmony search algorithm for solving structural optimization problems. The harmony memory considering rate and pitch adjusting rate are conceived as the two main parameters of the technique for generating new solution vectors. In the standard implementation of the technique appropriate constant values are assigned to these parameters following a sensitivity analysis for each problem considered. The success of the optimization process is directly related on a chosen parameter value set. The adaptive harmony search algorithm proposed here incorporates a new approach for adjusting these parameters automatically during the search for the most efficient optimization process. The efficiency of the proposed algorithm is numerically investigated using two large-scale steel frameworks that are designed for minimum weight according to the provisions of ASD-AISC specification. The solutions obtained are compared with those of the standard algorithm as well as of the other metaheuristic search techniques. It is shown that the proposed algorithm improves performance of the technique and it renders unnecessary the initial selection of the harmony search parameters.