In this study, we explore the applicability of (the peak frequency of the site‐response transfer function) as a site‐effect parameter for sites in California. The study is motivated by the results of our companion paper (Hassani and Atkinson, 2017), in which we show that peak frequency is a primary site variable for central and eastern North America (CENA), providing important information that is missing if the time‐averaged shear‐wave velocity over the top 30 m () is used as the sole site parameter. Specifically, we found that significant site‐response trends in peak frequency remain after removing site effects parameterized by , for sites in CENA. We therefore modeled site response using both peak frequency and and found this led to reduced variability. This led us to wonder if similar trends are present in California, where the use of as the primary site‐response variable is common practice. Might variability in site response also be reduced in California using the combination of and ? We use the Next Generation Attenuation West2 ground‐motion database to show that the answer to this question is “Yes,” but we also note that the effect of is weaker in California than in CENA. By following the same methodology as used in the companion paper for CENA and incorporating as an additional site parameter, we can reduce the site‐to‐site variability component of ground motion (by 5% on average). We conclude that including as an additional site‐response variable is warranted in California, in addition to its applicability for sites in CENA.