Since S-wave velocity of the subsurface is an important parameter in near surface applications, many studies have been conducted for its estimation. Among the various methods that use surface waves or body waves, Rayleigh wave inversion is the most popular. In practice, the densities and P-wave velocities of different layers are usually assumed to be known to avoid ill-posed problems, as they have less influence on the dispersion curves. However, improper assignment of these two groups of parameters leads to inaccurate estimation of the S-wave velocity profile. In order to address this problem, the all-parameters Rayleigh wave inversion strategy is proposed in which the S-wave velocities, layer thicknesses, densities and P-wave velocities of different layers are included as the unknown parameters for inversion. Meanwhile, the transitional Markov Chain Monte Carlo (TMCMC) algorithm is applied for the implementation of all-parameters Rayleigh wave inversion. One simulated example and two real-test applications are demonstrated to verify the capability of the proposed method in the estimation of the S-wave velocity profile, the densities and the P-wave velocities. Furthermore, it is verified that the proposed method achieved more accurate S-wave velocity profile estimation than the traditional approach.