The acoustic impedance of different fluids was indirectly measured by analysing the reverberation of a pulsed ultrasonic transducer showing a design as applied in contemporary process instrumentation. The set of observed fluids comprises air, methanol, ethanol, toluene, water and glycerine, which are substances of extremely different acoustic properties. The fluid impedance was precisely gauged using statistical linear models. The model complexity as well as the required calibration effort was minimized including a priori knowledge about the electrical characteristic of the transducer near resonance. The predictive power of different linear model estimates in terms of ordinary least squares, principal component regression and partial least squares was compared. A compensation method was proposed correcting the impact of temperature variation on analysed data. Additionally measured transit time permits the precise estimation of the density of arbitrary homogeneous fluids of varying temperature using traditional transducer designs. Thereby the functional range of conventional ultrasonic devices in process industries can be extended without the need for constructive modifications.