English-medium instruction (henceforth called EMI) refers to the use of English as a language of instruction in settings in which English is not an official language. This phenomenon has started to attract attention from policy makers and researchers as English becomes increasingly used as a language of instruction in tertiary education. This study therefore focuses on students’ attitudes toward EMI in Norway and French-speaking Belgium through a mixed-methods approach, i.e. a questionnaire survey followed by semi-structured interviews. Attitudes are conceptualised as a form of habitus related to the perceived chance of linguistic profit, in the sense of Bourdieu (1977), conditioning students’ choices and behaviour towards EMI. These attitudes are tested against socio-affective variables related to second language acquisition to verify how both sets of variables correlate, as feelings of unease using a foreign language might impact on the attitudes to EMI and act as a barrier as regards the enrolment or the participation in EMI programmes. Furthermore, I study the potential influence of students’ socioeconomic status on their participation in and attitudes towards EMI, as such a type of instruction may be addressed to an elite due to the language requirements to access it. The results of this study show that, even though Bourdieu’s theories are not completely successful to define their attitudes, students are neutral to positive towards EMI in both Norway and French-speaking Belgium. The legitimacy of EMI seems primarily based on its role as a preparation for English-speaking workplaces and on the class composition, as the presence of international students and teachers makes English the necessary lingua franca. EMI is therefore described as being useful to students in the case of the former, and necessary in a globalised world in the case of the latter. Although students tend to consider themselves as second language users rather than learners, attitudes correlate moderately with the socio-affective variables involved in second language acquisition, showing that these aspects should be taken into account in education planning. As regards students’ socioeconomic status, there is no correlation with students’ attitudes. However, the study shows that, surprisingly, participation in EMI is associated with lower socioeconomic backgrounds, showing that EMI is far from being a discriminatory practice.