This paper reviews recent evidence on the impact of a teacher incentives program in Kenya. The results are based on a randomized trial, which removes many sources of bias that can arise in analyses of non-experimental data. One hundred schools in a rural area were randomly divided into 50 that participated in a teacher incentives program and 50 that served as controls. Students in the schools that were selected the teacher incentives programs had higher test test scores on exams linked to incentives during the time the program was in place, but test scores did not increase significantly on exams that were not linked to incentives, and test score gains on exams linked to incentives did not persist after the program ended. Teachers in the 50 program schools did conduct more test preparation sessions, and students in those schools were more likely to take exams. Overall, there is little evidence that the teacher incentives program increased student learning. (Contains 4 tables and 11 footnotes.)