This paper presents a characterization of an underwater positioning system based on surface nodes equipped with GPS and acoustic transducers. The positioning system calculates the coordinates of an underwater vehicle in one of the surface nodes or beacons, by the emission, detection, and reply of acoustic encoded signals. The characterization of the system has been performed by means of a statistical study, considering different numbers of beacons, beacons’ position and physical phenomena, such as noise, multipath, and mems. The error propagation caused by these phenomena and the geometrical configuration of the system has been quantitatively assessed in different positioning algorithms, based on trilateration and iterative procedures. The results show how the different phenomena affect the vehicle estimated position errors for the different positioning algorithms. In addition, the obtained errors inside the projected area of the beacons are ∼1 m or lower, rising to a few meters for the worst case scenario, showing the feasibility of the acoustic positioning system