This talk details the findings of a recent study (McKenzie et al., 2019) – employing a free classification instrument – to investigate 191 Thai university students’(mis)categorisations of the geographical origin of nine L1 and L2 users of English from stimulus speech. Analysis indicated participants were generally able to distinguish between native and non-native English speech, and this distinction was found to be the primary perceptual dimension underlying speaker provenance categorisations. It was also found that the social-psychological salience of the speech forms, rather than geographical proximity, was key in determining categorisation accuracy whereby, for instance, recognition rates for Thai, UK, US and Indian English speakers were substantially higher when compared to Vietnamese and Australian English speakers. Analysis of misidentification patterns showed a tendency for the Thai students to conflate Asian English speech forms, despite substantial phonological and phonetic differences between the English spoken in different Asian nations. A discussion is offered with regard to current language attitude theory, in light of the evidence that a) segmental features were largely responsible for provenance (mis)categorisations and b) speaker categorisation processes seemed to lead to social evaluation of the speakers’ perceived group membership more widely.