Best practice research in secondary science education begins with a noble goal: learn from the excellence of past instruction in order to help mold future instructional practice. On the surface, it seems that all people need to do is find excellent science teachers, find out what they do in the classroom, and recreate this in other classes around the globe. Such a simplistic view of educational practice typically fails because individual differences have been neglected. Seeking to recreate the product that someone else has developed is analogous to teaching to the test. A less myopic, more robust, vision clarifies the conditions by which best practice succeeds. If transformations in teaching practice are achieved by mimicking best practice performances, every participant who attends a National Science Teachers Association (NSTA) conference would be transformed each time they see an exemplary practitioner's presentation--the research tells otherwise. Specifically, professional development experiences need to be personalized and sustained to create significant, lasting classroom improvements. Short professional development experiences can be motivational in the short term but lack sufficient time to be internalized into the curriculum. Tens of billions of dollars are spent each year on improving teacher quality, yet the goal of providing high-quality instruction in each classroom is far from being reached. In fact, test scores are not gaining at a rate that will satisfy either the goal of Science for All Americans--Project 2061, or the tenets of No Child Left Behind (NCLB). So what is working? What is not working? What can be shared globally? In this article, experienced science teachers are compared to Presidential Awardee science teachers in an attempt to explain the process by which high-quality secondary science teaching practice develops. (Contains 8 tables and 4 endnotes.)