Learning communities can trace their origins to the 1920s and the experimental College at the University of Wisconsin, but the widespread use of learning communities to improve undergraduate education is a relatively new phenomenon. The early learning communities of the late 1980s and early 1990s consisted of groups of new freshmen co-enrolled in several courses, including a seminar intended to help students make connections across courses and disciplines. Today, freshman learning communities may soon be the norm on college campuses. The reason for the growing popularity of learning communities is simple: they work. Despite the broad mandate of learning communities in general, individual institutions must assess the impact of their particular programs in order to determine their effectiveness. This article presents four key research-based lessons that illustrate how and why learning communities work and suggest appropriate assessment strategies: (1) Effects of learning communities on students are indirect, not direct; (2) Nature of the learning community influences the outcomes that are observed; (3) Effects of learning communities are influenced by the characteristics of the participants; and (4) Effects of learning communities may vary among institutional types.