Online data collection allows for access to diverse populations. In the current study, we used online recruitment and data collection methods to obtain a corpus of read short stories, real English words, and nonwords from adult talkers representing three authentic regional dialects of American English and one novel accent. The authentic dialects are New England, Northern, and Southern American English and are each represented by 8–10 talkers, ranging in age from 22 to 75 years old. The novel accent was produced by five Spanish-English bilinguals with training in linguistics, who were asked to produce Spanish /o/ in an otherwise English segmental context. The four target varieties each contain one vowel pair of interest, in which the vowels within the pair are relatively more ambiguous than in the other varieties. Each talker produced one familiar short story (e.g., Goldilocks and the Three Bears) with 40 tokens of each vowel within the target pair for their dialect, as well as a set of real words and nonwords that represent both the target vowel pair for their dialect and the other three vowel pairs for comparison across dialects. All corpus materials are available to the scholarly community.