This article critically reviews the research literature on cognitive outcomes of service-learning over the past decade with an emphasis on how convincing the results are to faculty. Self-report measures produce the most consistent positive findings yet are one of the least persuasive measures to faculty. The use of problem-solving protocols shows promise in measuring both student knowledge and the complexity of student thinking but needs further development. Recent work in the learning sciences provides direction for future outcome research and suggests how service-learning will help to transform education.