Tests are powerful learning tools when students complete them alone, but in practice, testing sometimes manifests as a group exercise. Collaborative learning approaches are generally embraced by educators although research in this area is mixed. Two questions linger: What memory mechanisms are at play when students test in groups? And when is collaborative testing most likely to enhance learning above and beyond individual testing? This article reviews both laboratory and applied research on collaborative testing, adopting a cognitive perspective and focusing on long-term retention of information. We discuss the well-established memory benefits of completing tests alone. Next, we summarize laboratory experiments on collaborative testing and compare them to applied studies that examined collaborative tests in real classrooms. Suggestions are made for educational practice and future research in the laboratory and classroom. We advocate translational research as a means to strengthen the theory behind collaborative testing, empowering researchers and educators alike.