Traditional mathematical optimization methods, such as linear programming (LP), nonlinear programming (NLP) and dynamic programming (DP), often have drawbacks in practice. To overcome these drawbacks, various meta-heuristic optimization techniques have been developed. This study proposes a new meta- heuristic algorithm which mimics musical performance, and names as Harmony Search (HS). The HS algorithm is applied to five problems: a continuous function optimization, a hydrologic parameter calibration, a traveling salesman problem, an optimal design of water supply network, and an optimal expansion of water supply network. The comparison of solutions obtained by HS and other optimization techniques shows that the HS algorithm can find better solutions in less number of iterations. The new algorithm can solve continuous-variable problems as well as discrete- variable problems without unnecessary restrictions imposed by calculus-based algorithms.