The single best predictor of who will be a great teacher next year is who was a great teacher this year. Some great teachers are seasoned veterans, while some are new to the field. Experience matters, of course, particularly in the early years of a teaching career, but it's no guarantee of teacher effectiveness. There is a wide variation in performance, even among teachers with the same amount of experience. This concept--that a great teacher cannot be identified on the basis of a résumé--should be liberating. For some reason however, it's largely ignored in the education field. Instead, without worrying too much about whether requirements will make teachers more effective, each state imposes myriad conditions on potential teachers and the programs they must attend to be eligible to teach. The result is requirements that amount to little more than barriers-new ways to keep potential teachers out of schools. This paper attempts to trace what is known about identifying and training successful teachers. It starts with a sobering conclusion: Although candidates spend billions of dollars and thousands of hours on teacher preparation courses, there still is no body of evidence justifying those requirements. Nor does anyone know how to measure and define a successful teacher training program. This paper suggests that policymakers should invest much more time and resources into learning about the science of teaching and how individual teachers develop their skills; and in the meantime, perhaps policies should reflect the fact that far more is known about a teacher after he or she enters the classroom, than beforehand. The authors propose and discuss four strategies for ensuring that schools and students have access to the best teachers possible.