Reviewing code you don't understand is an important aspect of bringing together different minds to build working software. While best practices exist for conducting code reviews, situations where the reviewer doesn't fully comprehend the code can be daunting. To address this, it's essential to identify high-risk areas in the diff, focus on tests that provide context, state assumptions, and leave comments as you work through the diff to express your understanding and tentatively ask questions. Additionally, reviewing at the level of understanding you have, manually verifying behavior, and providing constructive feedback can be valuable even when the implementation is complex or domain-specific. By adopting these strategies, reviewers can feel more confident in their ability to provide helpful feedback while acknowledging areas where they may not fully understand the code.