In response to the large computational load and extended processing time associated with the Monte Carlo method in the robustness verification of helicopter flight control systems, this paper proposes an Unscented Transform method for such verification. This method employs Particle Swarm Optimization for determining the uncertain parameters. Using the helicopter"s yaw channel as a case study, the Unscented Transform method"s performance in robustness verification was compared with that of the Monte Carlo method. Simulation results show that the Unscented Transform method significantly enhances computational efficiency while maintaining comparable accuracy to the Monte Carlo method in the robustness verification of helicopter flight control systems. By adjusting parameter values to modulate the distance between sampling points and the sample center, this method is also suitable for the robustness verification of high-dimensional helicopter flight control systems.