The study describes a pedagogic adaptation of the matched guise technique with the aim to raise linguistic self-awareness of L2 accentedness stereotyping effects among Swedish pre-service teachers. In the experiment, 290 students attending teacher training programs were exposed to one of two matched guises, representing either L1 accented Swedish, or L2 accented Swedish. Both guises were based on the same recording, but the L2 accented version had been digitally manipulated using cut-and-paste techniques in order to replicate certain vowel sounds (the [u:]-sound in particular) associated with low-prestige Swedish L2 accentedness. The findings from this experiment were then used as starting point for language awareness raising activities. Our overall results show that the L2 accented manipulated recording was evaluated more favourably than the original L1 accented recording on all investigated variables. One proposed explanation is that respondents were inadvertently influenced by so-called shifting standards effects, i.e. lower standards/expectations are being used as reference points when evaluating the L2 accented recording. This tendency, however, seemed to be less apparent among respondents with bi/multilingual linguistic identities. Following debriefing discussions based on the experiment findings, there were clear indications that respondents did become more aware of inadvertent linguistic stereotyping by participating in the activities.