In the U.S., differences between ethnolectal varieties such as Latinx English and African American English have been well-described for decades, especially in the realm of segmental phonological variation. However, few studies have examined patterns of voice quality features across ethnolectal varieties of U.S. English. The current study examines acoustic correlates of creakiness, modalness and breathiness (H1-H2, HNR, and CPP values) in the speech of 24 Southern California high school students: 17 Latinx and 7 Black American. Results of multiple regression analyses reveal significant differences in phrase-final voice quality, with Latinx students displaying patterns that are breathier and Black students displaying ones that are more modal or creakier. These results may be due to different strategies across ethnolects for accomplishing acoustic cues to phrase-final position. They can also reflect different levels of participation in the ongoing increase in the use of creaky voice among young speakers of Mainstream American English. This study acts as a first step towards understanding how racialized communities in the U.S. may participate in sociophonetic changes in progress that interact with the making of prosodic meaning.