The paper explores the ways in which research and practice have shaped how we think about the proper size of the high school. Since 1970, essentially all research tends to favor the creation of small high schools. Four forces seem to have come into play much more strongly, often in concert with each other, since our conception of the high school was established; the information age, the emergence of an adolescent culture, the studentsi rights movement, and our changing views of the proper functioning of organizations all threaten to render the large, comprehensive high school obsolete. Two reform responses to these conditions have emerged: break existing big high schools up into small schools within schools (SWASs) and create new, small high schools. The two alternatives represent very different models of schooling with very different cultures. Successful examples of the SWAS approach are rare, perhaps non-existent. Five types of error are often committed in designing them: errors of size, of continuity, of autonomy, of time, and of control. Creating new, small high schools is an effective approach but their effectiveness begins to diminish when they get much larger than 200 students. SWASs also seem to hit a size barrier when they try to get much smaller than 400 students. This no-manis-land of high school size is discussed and its implications for reform efforts are considered. The analysis attempts to explain why so little of our widespread effort to reform the high school has been successful and it suggests remedies, uncomfortable as they may be, that hold the promise of rectifying the situation.