Impact load measurements were carried out in a high-speed cavitation loop by means of a conventional pressure sensor flush-mounted in the region of closure of the cavity where maximum damage was observed. The sensor was dynamically calibrated by the ball drop test technique. Pressure pulse amplitudes were measured at different velocities and constant cavitation number and cavity length. It was found that pressure pulse height spectra follow a simple exponential law, which depends upon two parameters interpreted as a reference peak rate and a reference load. By exploring the dependence of both parameters on flow velocity, it was possible to show that the various histograms measured at different velocities can be reduced to a unique non-dimensional one and derive scaling laws, which enable to transpose results from one velocity to another. The measured values of impact loads are compared to similar data in the literature, and the impact load spectra are discussed with respect to pitting test results available from a previous investigation. It is concluded that an uncertainty remains on the measured values of impact loads and that a special effort should be made to compare quantitatively pitting test results and impact load measurements. To evaluate the coherence of both sets of data with each other, it is suggested to introduce two-dimensional histograms of impact loads by considering the size of the impacted area in addition to the measured impact load amplitude. It is conjectured that the combination of impact load measurements and pitting test measurements should allow the determination of such two-dimensional histograms, which are an essential input for analyzing the material response and computing the progression of erosion with exposure time. [DOI: 10.1115/1.4005342]