Rayleigh waves have high amplitude, low frequency, and low velocity, which are treated as strong noise to be attenuated in refl ected seismic surveys. This study addresses how to identify useful shear wave velocity profile and stratigraphic information from Rayleigh waves. We choose the Firefl y algorithm for inversion of surface waves. The Firefl y algorithm, a new type of particle swarm optimization, has the advantages of being robust, highly effective, and allows global searching. This algorithm is feasible and has advantages for use in Rayleigh wave inversion with both synthetic models andfi eld data. The results show that the Firefl y algorithm, which is a robust and practical method, can achieve nonlinear inversion of surface waves with high resolution.