Because English is widely used as a lingua franca, language testers have started to consider the introduction of non-native accents into English listening tests. This study investigates how accents influence test-takers' performance, and also elicits test-takers' subjective perception of accents. Eighty adolescent L1-Mandarin test takers were divided into four groups of equal proficiency, with each group listening to one accented version of the same English listening test. The test input was delivered in Australian, Spanish, Vietnamese, and Mandarin-accented English varieties with tasks measuring lexical and propositional comprehension and consisting of picture selection, true/false, and gap completion items. Test-takers' perceptions of accent familiarity, comprehensibility, and their attitudes were also measured. Results indicate that the test takers who received Mandarin-accented input performed best, lending support to a shared-L1 effect, with the strongest advantage for lexical comprehension. No significant difference was observed in test scores among the groups exposed to non-Mandarin accents. Findings also reveal that the type of accent was not significantly related to test-takers' attitude toward it. The central implication from this study is that there is potential for the inclusion of non-native accents into listening tests for adolescent learners if the shared-L1 effect can be addressed.