Background: Although reporting on the health care-setting level of continuance or discontinuance of an intervention once a trial is completed has been recommended, such "real-world" diffusion studies are rare. The present example was made possible by funding to explore opportunities for post-trial implementation of an innovative health counseling intervention for cardiovascular prevention in The Netherlands. Methods: Between 2001 and 2004, in a longitudinal case study, we compared two healthcare settings: a cardiology outpatient clinic and general practices. Rogers' diffusion of innovations theory served as the theoretical background. Information was extracted from minutes of meetings and informal conversations with health counselors, and checked by the project manager. Additional data were collected from physicians with a short questionnaire. Results: Implementation of the health counseling intervention was successful in the cardiology outpatient clinic, but was unsuccessful in the general practices. Success was related to a centralized diffusion system, stronger "change agent" efforts, avoidance of post-trial interruption of service delivery, easily achievable "reinventions," and positive physician perceptions of the service (i.e., not complex and compatible with current practice routines). Support came from changes in the organization of care that created opportunities for, instead of competition with, the innovative service. However, coincidental events may also have played a part. Conclusions: Our findings confirm the importance of most theoretically predicted individual and organizational diffusion variables. This implies that the implementation of innovative healthcare services requires attention at both levels.