Good child care quality is associated with good child outcomes, but high-quality care is rare. What level of quality is necessary to achieve good outcomes? Is there a minimum level of quality that must be delivered in order to attain good outcomes? Is there a level of quality beyond which there are diminishing returns? To address such questions, the goal of this project is to understand the shape of the relationship between child outcomes and a measure of teacher-child interactions (the Classroom Assessment Scoring System, or CLASS) that is increasingly used in program improvement and in high-stakes applications such as quality rating and improvement systems and Head Start designated renewal. The distinguishing feature of this project is the use of contemporary quantitative techniques (shape constrained additive models) to rigorously estimate the shape of the relationship.