Program sustainability is an ongoing concern for most people in health promotion. However, the current notion of sustainability in organizations, namely routinization, needs refinement. This article examines organizational routines. In so doing, it refines the notion of sustainability and the assessment of routines. Drawing on the organizational literature, a routinized program is defined by the presence of routinized activities, meaning that these activities exhibit four characteristics of organizational routines: memory, adaptation, values and rules. To answer the question of how these characteristics are useful, we conducted an empirical study of the routinization of the Quebec Heart Health Demonstration Project in five community health centers. Our method consisted of a multiple-case study. We observed project activities in each center in 2000. The data came from documents and interviews with project actors. Our results show that, in one of the centers, no resources had been officially committed to project activities. Even so, the actors continued some activities on an informal basis. In another center, the activities satisfied three of the four routine characteristics. In the three others, activities satisfied all of the characteristics. These results suggest focusing the study of program sustainability on the routinization of activities resulting from it. They indicate four distinct degrees of sustainability: (1) the absence of sustainability; no program activity is continued; (2) precarious sustainability; some residual activities are pursued, at least unofficially; (3) weak sustainability; the program produces some official activities that are not routinized; and (4) sustainability through routinization; routinized activities result from the program.