AbstractArias intensity (IA) and peak ground acceleration (PGA) have been widely used as measures of the intensity of strong ground motions. This study employs a large set of data consisting of 7034 horizontal and 3474 vertical strong-motion records from 173 worldwide earthquakes (Mw 4.3–7.9) to refine the relationship between IA and PGA and investigates its potential dependence on variables such as earthquake magnitude (Mw), local site condition (VS30), focal mechanism, and fault distance. The dataset from the 2011 Mw 9.0 Tohoku earthquake and the 2013 Mw 6.8 Lushan earthquake is used to demonstrate the usability and necessity of our models. The results reveal that the logarithm of IA increases linearly with the increase of the logarithm of PGA and Mw and the decrease of the logarithm of VS30, and this kind of correlation is not significantly affected by focal mechanism and fault distance. New global empirical relationships for both horizontal and vertical components are developed to estimate IA as a function of PGA, Mw, and VS30. The resulting correlation equations presented in this article represent a significant advancement by incorporating such important features as magnitude and VS30 and enable an improved way of estimating IA from PGA.