Raman spectroscopy can be employed to measure the chemical composition of a sample, which can in turn be used to extract biological information. The aim of this paper is to introduce an efficient simulation technique for Raman spectroscopy in turbid (scattering) media taking into account relevant detector parameters and the sampling volume. We simulate the process of photon motion in turbid media by means of the Monte Carlo (MC) method. The numerical simulation of Raman scattering consists of two stages: calculation of the photon fluence at each point of the medium and subsequent generation of the corresponding amount of Raman photons at each point. The developed model allows simulation of both confocal and optical fiber probe Raman setups. In more detail, the model efficiently simulates Raman signals for different single and multi-layer phantoms and geometries, including focused and collimated (i.e., the fiber-based case) excitation laser beams as well as different values for the numerical aperture and the excitation beam radius. In the future, our results offer the potential to improve the design of Raman systems for in vivo applications in biomedical research.