Although L2 listening comprehension is an important component of language learning, it is less researched than more well-established aspects, such as L2 reading and writing (Goh, 1998; 2017). Recent research (e.g., Rost, 2011; Vandergrift & Goh, 2012) suggests that listening comprehension is in fact a complicated, dynamic, and integrative process, involving the L2 listeners’ active mental activities at a particular point of listening. Listeners’ successful comprehension, therefore, may be subject to the negative influence of the affective factors, such as listening anxiety, test anxiety, background knowledge and etc. This study aims to test the mediating effect of listening metacognitive awareness between listening anxiety and listening test score, as well as between test anxiety and listening test score, among a sample of 402 Chinese English as a Foreign Language test takers. Participants filled in the questionnaires in a self-reported manner and structural equation modeling was used to analyze the data. The study validated and explored this mediation model put forward by the author of the present study based on the existing literature. To this end, confirmatory factor analysis (CFA) and structural equation modeling (SEM) were used to validate the constructs of listening anxiety, test anxiety, and listening metacognitive awareness, as well as to test the indirect and direct effects of the structural pathways. In addition, multi-group analysis (high and low listening proficiency group) was run to generate alternative explanations. Results showed that listening metacognitive awareness mediates the relationship between listening anxiety and listening test score, as well as the relationship between test anxiety and listening test score. In addition, listening metacognitive awareness mediates the relationship between listening anxiety and listening test score in a low listening proficiency groups but not in a high listening proficiency groups. Finally, listening metacognitive awareness mediates the relationship between test anxiety and listening test score in both groups. It is believed that the findings of this study have implications for college English teachers' L2 listening instruction methods and offers ways to alleviate English as a foreign language (EFL) learners’ listening anxiety and test anxiety. Furthermore, the results are important for EFL learners to improve listening metacognitive awareness. Methodologically, this study shows how SEM is conducted to enable statistical analysis that is not supported by other commonly used statistical techniques, such as linear multiple regression, particularly by demonstrating how the mediating effect is tested by using bootstrapping.