Controlled educational choice, defined as parental or student choice of educational programs within the public school system that is regulated for public purposes, has become a leading proposal for educational reform. This paper reviews evidence from three working models of choice: the Minnesota Postsecondary Enrollment Options program (PEO), the Washington State Educational Clinics program, and the Community District 4 Alternative School Choice program in New York City. Each program was characterized by early opposition, followed by relatively smooth implementation. All three programs show evidence of positive effects, though none has demonstrated the extent to which choice enhances student or school performance. In none of the models is choice available to, or exercised by, all the system's intended clients. Introducing controlled choice does not necessarily result in a deregulated system stressing parent and school association around common purposes. In fact, all three choice models are heavily regulated on either the demand side, supply side, or both; this regulation is largely responsible for their effects. The programs demonstrate a wide variety of design issues that interested policy makers must confront, including the degree of regulation on supply and demand, the possible negative effects of choice systems benefiting active or inactive choosers, and the limited effect of choice systems on the broader systems in which they operate. (9 references) (Author/MLH)