Event completion describes the finding that observers often report having seen some visual information that was missing (e.g., the moment of ball contact) if it is causally implied (e.g., by seeing a ball flying). We investigate whether the accuracy of observers’ decisions depends not only on the precision of their mental representation of the event, but also on their confidence in their cognitive processes. While confidence is often measured alongside accuracy in event perception research, to our knowledge, the concepts of accuracy and confidence were never put into relation. Leveraging Signal-Detection Theory measures, we calculate metacognitive sensitivity (meta-d'), a bias free measurement of the confidence-accuracy relationship, and apply it to the event completion paradigm. We will present both a re-analysis of existing data from the event completion paradigm (Huff et al., 2017; Papenmeier et al., 2019) using meta-d' and the results of a new study (pre-registered: https://aspredicted.org/BK7_XH9) using the event completion paradigm that is specifically designed to ensure accurate measurement of meta-d’.Participants watch soccer video clips and indicate whether they have seen (including confidence rating) a critical moment of ball contact that is either causally implied (e.g., the ball flying; causal continuation) or not (e.g., cheering supporters; non-causal continuation). In half of the trials, we omit this critical moment of ball contact. We expect that detection accuracy and metacognitive sensitivity is lower in the causal than in the non-causal condition. Our research will contribute to a better understanding of peoples’ insight into the accuracy of their judgments during event perception.