Abstract. Post's correspondence problem (PCP) is a classic undecidable problem. Its theoretical unbounded search space makes it hard to judge whether a PCP instance has a solution, and to find the solutions if they exist. In this paper, we describe new application-dependent methods used to efficiently find optimal solutions to individual instances, and to identify instances with no solution. We also provide strategies to create hard PCP instances with long optimal solutions. These methods are practical approaches to this theoretical problem, and the experimental results present new insights into PCP and show the remarkable improvement achieved by the incorporation of knowledge into general search algorithms in the domain of PCP. 1