I work on a team that runs following a high trust and domain ownership model. When I joined our process was "write some code, test it, be confident in the code, commit it to trunk". A number of years ago we worked our way through "Clean Code" in a reading group and decided to try out code reviews. Each week one of us would present our recent work and the team would tear it apart line by line, and in the process we developed a better understanding of our group philosophy around style and practices.

When we switched to Git we decided that we liked code reviews and wanted to make them part of merging to develop. Depending on the content the reviews are deeper or more superficial. Many times even when I review the code of somebody making changes in their own subject matter domain I have found bugs that simply needed another set of eyes to catch.

I like code reviews and in the context I use them find that they are a complement to trust, not a substitute.