The assessment of pragmatic skills in a foreign or second language (L2) is usually investigated with regard to language learners, but rarely with regard to non-native language instructors, who are simultaneously teachers and (advanced) learners of the L2. With regard to English as the target language, this is a true research gap, as nonnative English-speaking teachers (non-NESTs) constitute the majority of English teachers world-wide (Kamhi-Stein 2016). Addressing this research gap, this paper presents a modified replication of Bardovi-Harlig and Dörnyei’s (1998) renowned study on grammatical vs. pragmatic awareness, carried out with non-NEST candidates. While the original study asked the participants for a global indication of (in)appropriateness/ (in)correctness and to rate its severity, the participants in the present study were asked to identify the nature of the violation and to suggest a repair. Inspired by Pfingsthorn and Flöck (2017), the data was analyzed by means of Signal Detection Theory with regard to Hits, Misses, False Alarms and Correct Rejections to gain more detailed insights into the participants’ metalinguistic perceptions. In addition, the study investigated the rate of successful repairs, showing that correct problem identification cannot necessarily be equated with adequate repair abilities. Implications for research, language teaching and language teacher education are derived.