English vowel formants undergo phonetic enhancement under prosodic prominence. However, vowel classes differ in the extent and dimensions of enhancement in ways that remain poorly understood. This study investigates differences in the effect of prosodic prominence on the production of eight English vowels. Thirty speakers each produced 48 repetitions of nonce words containing critical vowels. To elicit prominence, speakers produced short phrases that either repeated after a model talker or corrected the talker’s use of one word. Bayesian multivariate mixed-effects models were used to assess the effect of prominence and vowel on F1 and F2. Results suggest an inverse relationship between vowel height and effect of focus on F1, as well as a trade-off between the use of F1 and F2 to signal prominence. For the lowest vowels examined, /ɛ/, /ʌ/ and /a/, prominence was associated with raised F1 and no accompanying movement of F2. Conversely, the high front vowels /i/ and /e/ peripheralized along F2 under prominence without moving along the height dimension. Vowels intermediate to these showed movement along both dimensions. We discuss these results in terms of our understanding of the relationship between the expression of prominence and the phonological specification of vowels.