It is well established that teachers are the most important in-school factorin determining student outcomes. However, to date there is scant robustquantitative research demonstrating that teacher training programs can havelasting impacts on student test scores. To address this gap, weconduct andevaluate a teacher peer-to-peer observation and feedback program underRandomized Control Trial (RCT) conditions. Half of 181 volunteer primaryschools in England were randomly selected to participate in the two yearprogram. We find that students of treated teachers perform no better onnational tests a year after the program ended. The absence of externalobservers and incentives in our program may explain the contrast of theseresults with the small body of work which shows a positive influence ofteacher observation and feedback on pupil outcomes.