Beyond the Tennessee experiment, a wide range of simple to sophisticated analyses controlling for family background and teacher differences shows no systematic relationship between class size and student performance.