This paper traces the history of graduated licensing, starting about the point in time when Pat Waller's paper on the genesis of the concept ends, and examines the extent to which graduated licensing has produced reductions in collisions. It concludes with some general observations about future research needs, anticipating several of the papers that follow. The evolution of graduated licensing is chronicled, beginning with the early and largely unsuccessful efforts to introduce it in the United States in the late 1970s, through the pioneering efforts in New Zealand, which resulted in the first truly graduated system in 1987, to Canada where the program was introduced 7 years later, to the United States where it has flourished in more recent years. This 25-year history lesson hopefully creates an appreciation for the somewhat torturous journey that graduated licensing has experienced in achieving acceptance among the public and policy-makers-a journey that is not yet over, as subsequent papers in the symposium will show. The proliferation of graduated licensing in recent years is a mixed blessing-the wider adoption of graduated licensing has been a very positive development, but the programs that have evolved are anything but homogeneous in structure or content. Although this is often necessary for various reasons, it is worrisome that some programs are graduated licensing in name only. This suggests that future efforts to promote graduated licensing must emphasize adherence to the fundamental risk reduction and multistage principles on which the concept is based. The paper also considers the extent to which graduated licensing achieves its objective of reducing collisions among those covered by the program. Understandably, most jurisdictions would not introduce graduated licensing until it was shown to be effective and this, to some extent, slowed the process of implementation. The obvious irony is that it could not be shown to be effective until it was introduced. Fortunately, as history demonstrates, some jurisdictions were prepared to try the system based on its very sound empirical rationale. And, their confidence has been rewarded. A growing body of research, which shows that graduated licensing has been associated with significant and substantial reductions in collisions, is briefly described. The paper concludes with some general observations designed to anticipate the papers that follow. First, it outlines questions that still remain unanswered about graduated licensing-why or how it works, with whom it works, and what features are most effective. Precise and unambiguous answers to these questions are essential for the design of a system that maximizes the potential for reducing collisions, injuries, and deaths. Second, it signals a note of caution on the limits of graduated licensing-it is important to recognize just how effective and beneficial this program is; it is equally important to recognize that it is not the sole panacea for the problem of collisions involving new drivers.