Soil porosity is a state parameter of fundamental importance for several geotechnical problems. Geophysical testing provide appealing strategies for the determination of soil porosity, as several geophysical parameters are directly related to soil porosity. In particular the theory of wave propagation in saturated porous media, developed by Biot in the 1950s, allows the determination of soil porosity from the measured velocity of propagation of compressional and shear waves. A formal assessment of the reliability of the estimated porosity values is of primary importance to evaluate the applicability of this approach to solve practical geotechnical problems. In this paper the propagation of measurement uncertainties on the estimated values of soil porosity is theoretically evaluated. Moreover, experimental data of multiple acquisitions of cross-hole tests are considered. Data collected by different operators are also used to assess the confidence interval associated to different equipment, acquisition practices and testing methodology.