This study examines the rate and quality of L+H* contours in five varieties of American English. Reading passage data from 30 female participants (10 Jewish English (JE), 10 African American English (AAE) and 10 Appalachian English (ApEng)) was coded using MAE-ToBI conventions. Mixed- effects modeling was used to compare the number of instances of L+H* and H*/!H* contours, and peak contour height, slope, and peak offset of the L+H* contours. ApEng speakers use the highest number of L+H* contours in their speech. JE speakers use fewer L+H* contours than ApEng speakers, but more than the AAE speakers. The phonetic implementation of the contour was also examined. ApE and JE speakers have higher peaks, wider rise spans, and steeper rises than AAE speakers, in parallel with the results for rate of use of L+H*. When compared to data for white speakers of Southern and Midland English (data from  ), the three groups of interest all use a higher proportion of L+H* contours. These results represent an important theoretical contribution by demonstrating that suprasegmental features are ethnolinguistically and regionally conditioned by rate of use and different realizations, in a manner similar to what has been previously observed for segmental phonological features.