A controlled experiment measured native Hong Kong Cantonese speakers’ perceptions of the quality of three different simultaneous interpretations (SIs) into Cantonese. The SIs differed only in the interpreters’ accents, native in one case and non-native in the other two. Participants were randomly assigned to one of three groups, listening to the following versions of the SI: (1) native-accented Cantonese (control group); (2) Mandarin-accented Cantonese; (3) English-accented Cantonese. To motivate participants to listen attentively, they were told beforehand that they would take a comprehension test before assessing the quality of the SI delivery. Ten questionnaire items measuring SI quality perception were analyzed quantitatively, while any additional comments were examined qualitatively. Overall, respondents in the two groups who listened to SI delivered with a non-native accent assigned lower quality ratings on all survey items than did respondents listening to native-accented Cantonese SI. Qualitative data suggest three possible explanations for the low ratings of the two SIs with a non-native accent: (1) extra listening effort was required to understand these SIs; (2) negative stereotypes were triggered by the interpreters’ non-native accents; (3) feelings of insecurity or threat were prompted by the non-native interpreters’ ability to perform SI into Cantonese.