There is both less and more to mastery learning (ML) than meets the eye. Less because mastery learning is not based on a model of school learning, and more because it is the most optimistic statement we have about the power of education. The notions of setting achievement standards and letting time for completion vary, of using criterion referenced tests, of the motivational value of success, of individualizing instruction according to student traits, and of formative evaluation and feedback are not unique to ML but what is unique to ML is the hypothesis that using these notions together can transform slow learners into faster learners. Recent state laws require mastery or minimum standards, but not that ML strategies be adopted. Questions should be answered about three aspects of ML: cost efficiency, equity, and externalities. ML will be characterized by rising costs as the target mastery level is raised. Equity refers to fairness in the distribution of resources to students, since ML calls for allocation of resources in roughly inverse proportion to prior achievement. An externality is a consequence of an action which affects others not party to the action. Marginal externalities should equal or exceed the marignal cost of obtaining them. (Author/CTM)