Unmanned aerial vehicles (UAVs) are favored by the industry to search and locate lost persons in mountains and trapped persons in earthquakes, fires and other disasters because it is not limited by the obstructions on the ground. Currently, however, a UAV always searches and locates targets along a fixed flight path, which consumes more time and has lower accuracy. This kind of method can only provide a rough position estimation. GuideLoc takes the UAV’s GPS coordinates as the location information of a target and the genetic algorithm (GA) is used for path planning in order to shorten the flight path to improve the search efficiency and obtain a good result. But its performance still has room for improvement. In this paper, the path optimization algorithm used in GuideLoc is further discussed and studied, and then a method, 2-OptACO, is proposed. The method uses the 2-optimization (2-opt) algorithm to improve the ant colony optimization algorithm (ACO) and is applied to optimize the UAV’s path for search and rescue. The simulation results show that the 2-OptACO method has a faster convergence rate than the GA and ACO. It can obtain a better global optimal solution.