The seismic inversion method based on rock physics constraints aims to deal with the challenge of accurately predicting physical properties of complex tight sandstone reservoirs, which often exhibit strong ambiguity. This ambiguity makes it difficult to characterize the reservoir and accurately select high-quality ones. By employing theoretical rock physics modeling and a Bayesian probability inversion framework, along with pre-stack seismic data and well logging data, a method is proposed to predict the physical properties of complex tight sandstone reservoirs through seismic analysis. This approach allows for a detailed characterization of these reservoirs and provides technical support for exploring such complex formations. Because complex tight sandstone reservoirs typically have low porosity, low permeability, and intricate pore structures, assuming a uniform pore structure makes traditional rock physics modeling and reservoir parameter seismic inversion methods inadequate. Hence, this paper derives seismic reflection coefficient equations for complex tight sandstone reservoirs, considering parameters like porosity, water saturation, and pore aspect ratio. Using these equations, a Bayesian pre-stack seismic inversion method is developed, which directly estimates the physical parameters and pore geometry of complex tight sandstone reservoirs using Bayesian linear theory. Results indicate that this method not only provides more accurate predictions of rock physics parameters but also yields physical property estimates that align better with actual wellbore measurements. This demonstrates the method's effectiveness and precision in predicting physical properties of complex tight sandstone reservoirs using seismic data.