The most recent attempt by educators to emulate the "sound principles" and methods of business and science is to become data-driven. The leaders in a data-driven school are able to demonstrate how some number, preferably scores on standardized tests, has moved upward as a result of some program they have initiated. Data-driven schools also possess the capacity to quickly access individual profiles of student performance on a wide variety of testing instruments. A sophisticated data-driven school provides teachers with lesson plans and instruction activities that will remediate the "diagnosed" learning deficiencies of each student in the school. But are data-driven schools all they are cracked up to be? In this article, the author discusses the seven myths of data-driven schools: (1) the numbers are accurate; (2) we are looking at the right data; (3) numbers can accurately describe the social phenomena; (4) test scores should always go up; (5) even if the numbers were right, we would do the right thing; (6) programs teach; (7) illusion of control. He asserts that school leaders need to stop looking at the data, no matter how beguiling its call, and start looking in the classrooms.