This paper offers guidelines for policy makers who are responsible for establishing incentive programs for teachers. It provides a framework for: (1) identifying performance-based incentive systems; (2) examining their underlying assumptions; (3) classifying program operations (types of incentives, methods of reward delivery, timing and types of performance standards and measures); and (4) checking that a program as implemented is ready for evaluation. The prospects for learning from incentive experiments are discussed. (JD)