This study quantified the microstructure of polyurethane foams and elucidated its relationship to fluid-acoustic parameters. The complex morphology derived from the three-dimensional images obtained by micro-computed tomography was analyzed using digital image processing and represented by a pore network model (PNM) and a distance map model. The PNM describes the fluid phase of a porous medium with equivalent spherical pores and circular throats, whereas the distance map model describes the solid phase with the average frame thickness. The porous materials were then modeled by six representative microstructural parameters that describe the geometry and topology of the fluid and solid phases. These parameters were pore radius, throat radius, distance between adjacent pores, coordination number, pore inclination angle, and frame thickness. Semi-phenomenological and empirical approaches were proposed to relate the microstructural properties to the fluid-acoustic parameters. These models effectively described the acoustic parameters and sound absorption performance of six different polyurethane foams. Since the representative microstructural parameters were obtained from small sample volumes of a heterogeneous material, notable variations were observed across different regions of the sample. Hence, this study quantified the effect of the uncertainty in each microstructural parameter on the resulting acoustic parameters using global sensitivity analysis.