This data was collected to answer the question of whether infants attend to linguistic and indexical information in vowel perception. Australian-English-learning and Dutch-learning 12-month-olds were played strings of vowel tokens from a central speaker. To measure their attention to the auditory stimulus, a Tobii eyetracker sampling at 120 Hz tracked their attention to a monitor in the vicinity of the speaker. Infants were first familiarized to eight 13-second trials containing repetitions of two tokens of the vowel /ɪ/ spoken by a female speaker of North Holland Dutch (trials Fam1 - Fam8). Infants were then presented with two blocks of three different test trial types, presented in random order within block: NoChange, which presented the same two tokens as in familiarization; VowelChange, in which infants heard the same vowel /ɪ/ alternated with tokens of the vowel /ɛ/ produced by the same speaker; and the IndexicalChange trial, which was either a Speaker Change, meaning that infants heard the same vowel /ɪ/ spoken by the familiarized speaker, alternated with tokens of the vowel /ɪ/ spoken by another female speaker of North Holland Dutch, or an Accent Change, in which tokens of the same vowel /ɪ/ spoken by the familiarized speaker were alternated with tokens of the vowel /ɪ/ spoken by a female speaker of another accent of Dutch, East Flemish Dutch. Data for each trial are infants' cumulative looking time in ms for each 13-second trial. If infants attended to the linguistic or indexical change in auditory stimulus, we would expect increased looking to that trial relative to the NoChange trial.