Abstract: In order to obtain the physical and geoacoustic properties of marine sediments, an inverse method using reflection loss of different grazing angles is presented. The reflection loss is calculated according to the reflection model of effective density fluid approximation. A two-step hybrid optimization algorithm combining differential evolution and particle swarm optimization along with Bayesian inversion are employed in estimation of porosity, mean grain size, mass density and bulk modulus of grains, respectively. Based on above physical parameters, geoacoustic parameters, including sound speed and attenuation, are further calculated. Numerical simulations show that all the parameters can be well estimated with the exception of bulk modulus of grains. In particular, this indirect inverse method for bottom geoacoustic parameters performs high accuracy and strong robustness. The relative errors are 0.092% and 17%, respectively. Finally, measured reflection loss data of sandy sediments at the bottom of a water tank is analyzed, and the estimation value, uncertainty and correlation of each parameter are presented. And the availability of this inverse method is verified through comparison between inverse results and part of measured parameters