AbstractDeveloping a reliable, cost-effective liquefaction triggering procedure for characterizing the liquefaction potential of gravelly soils based on in situ penetration testing has always been a great challenge for geotechnical engineers and researchers. Typical correlations based on the standard penetration test (SPT) and the cone penetration test (CPT) are affected by large-size gravel particles, which can lead to erroneous results. The Becker Penetration Test, well known for gravelly soil characterization, is cost-prohibitive for routine projects and is not available in most of the world. With a cone diameter of 74 mm the Chinese dynamic cone penetration test (DPT) is superior to smaller penetrometers and can be economically performed with conventional drilling equipment. DPT has previously been directly correlated to field performance data, and probabilistic liquefaction resistance curves were developed based on one earthquake and geologic environment in China; however, the use of these data in other tectonic and geologic environments was not validated. In this study, 137 data points from 10 different earthquakes and different depositional environments in seven countries have been used to develop probabilistic liquefaction resistance curves. The data set was expanded by performing DPT soundings at sites around the world where gravelly soil did or did not liquefy in past earthquakes. Based on the expanded DPT database, a new set of magnitude-dependent probabilistic triggering curves has been developed using logistic regression analysis. The new triggering curves are better constrained by data and the spread between the 85% and 15% probability curves is reduced. Liquefaction resistance is shifted upward at lower DPT values. A new magnitude scaling factor (MSF) curve has also been developed specifically for gravel liquefaction which was found to be consistent with previous curves for sand.