<p>Face masks can impact processing a narrative in sign language, affecting several metacognitive dimensions of understanding (i.e., perceived effort, confidence and feeling of understanding). However, to what extent this is modulated by sign language proficiency remains an open question. Thirty deaf adults were administered a narrative in Italian Sign Language (LIS), signed in three different visual conditions: no mask, transparent mask, opaque mask. In addition, they completed a Sentence Reproduction Task in LIS (LIS-SRT) to measure their sign language proficiency. Results showed that some of the error types in the LIS-SRT have significant correlations with performance in the comprehension task, revealing external validity for this sentence repetition test. Crucially, while sign language proficiency did not modulate the impact of face-masks on narrative comprehension, the metacognitive outcomes were clearly influenced by the interaction between LIS proficiency and visual conditions. Skilled signers experienced less effort and higher levels of confidence and feeling of understanding, whereas these subjective experiences were greatly impacted by visual conditions in less proficient signers. We discuss these findings in reference to cognitive load, thus also extending the construct of “listening effort” to sign languages. In addition, we relate differences in cognitive load in skilled vs. less proficient signers to the ability to extract and process at the same time multiple types of linguistic elements from the flow of signs (multilinear processing of sign language).</p>