In this article, we present a method to obtain dispersion curves using a Bayesian approach. This type of probabilistic analysis explores all possible results, and the final output is not unique but is expressed as a range of solutions. First, we describe how to obtain dispersion curves using manual procedures; then, we present a method that uses a probabilistic approach. The main idea is to use previous knowledge of a shear‐wave velocity structure to obtain a prior dispersion curve while simultaneously computing the expectation function of the spectral amplitude for a range of focal mechanisms. Then, these dispersion curves and expectation functions are simultaneously compared with the observed data, and the selected peaks of the dispersion curve are chosen using a weighted probability decision. In these two stages, we used a probabilistic approach based on Markov chain Monte Carlo (MCMC) with the Metropolis–Hasting sampling algorithm. The entire process is a Bayesian inference; thus, it is updated whenever new information is available in such a way that the result becomes a prior.