This article analyses a case of second-dialect performance as an idealised instance of second-dialect acquisition, without mitigating factors such as access, analytical ability and motivation. It focuses on the Australian English and American English speech of three young Australian actors. An acoustic analysis of their short-vowel systems shows that they can successfully adapt to perform in an American English accent, but that their second-dialect system is less stable and more variable than their native system.

A foreign-accent rating experiment on the actors’ American English with American English judges shows that the actors on average are thought to sound slightly less American than the native American English-speaker controls. The discrepancy between the acoustic accuracy and listener acceptability may be explained by judges attending to different features from those included in the acoustic study.

This study of second-dialect performance shows what is maximally possible in second-dialect acquisition. Given the difference between the two measures of success, studies of second-dialect acquisition would benefit from including subjective measures in addition to acoustic accuracy.