This study explores categorization processes of people (identities) and language (linguistic varieties) in interactions between L1 (first language) and L2 (second language) speakers of Japanese and the language ideologies behind them. Utilizing Conversation Analysis (CA) in combination with Membership Categorization Analysis (MCA), the present study focuses on how participants apply these categories to self and other where identities and language ideologies emerge in the sequences of ordinary conversations. The study also illuminates how the participants react to such ideologies, which is rarely documented in previous studies of L2 Japanese interactions. It is controversial to use CA and MCA as methodologies for inquiries into ideology due to different epistemological and theoretical frameworks. Yet, joining the emerging trend of CA studies that address ideological issues, this study will also demonstrate the compatibility between them. Methodological integration of CA and MCA has been proposed since the 1970s, but has started to be adopted only recently. Because few studies employ this combination in the area of language ideologies, it serves as a novel analytic tool in this body of research. Thus, this study makes a methodological contribution to the study of language ideologies, illustrating the production of language ideologies and reactions to it as participants’ accomplishments.