While continuity of care becomes a more pressing issue given changes in the locus of medical care under reform initiatives, our ability to understand how socio-medical factors shape the continued use of medical care services is limited by current conceptual and measurement tools. In this paper, we reconsider the possibility of measuring continuity of care for individuals during the course of an episode of illness. We recognize the inherent heterogeneity in the notion of "continuity" and suggest that different scenarios produce breaks in the use and delivery of services. We lay out two crucial distinctions. The first targets the type of breaks in continuous care, distinguishing among provider, sector and system continuity. Second, we examine the source of provider continuity, separating breaks in health service use that result from the client's failure to remain in treatment and the health care system's failure to provide for continuous care. We suspect that these scenarios represent different social processes and may be related to varying social, medical and organization factors. We argue that attempts to measure continuity of care must be tailored to both a longitudinal account of an illness episode and the specific nature of health problems. Here, we describe one way to gather data and we present an interview-based module focused on contact with formal practitioners consulted for selfreported mental health problems. Using data from the second wave of the ongoing, longitudinal study, Mental Health Care Utilization Among Puerto Ricans Study (1993-1994, N -3,263), we provide preliminary analyses of the extent and nature of continuity of care among individuals who live in low income areas of the island. We find that this approach is not only possible but feasible given the availability of computer-based algorithms. Just over one third (35.1 percent) of individuals meet criteria for our definition of continuity of care. Most breaks in service use are client-initiated (85.2 percent). We conclude by discussing the role of social scientists in reformulating theories of health care contacts that capture the experience of illness in an era characterized by community-based, managed care.Please address all communications to the first or second authors. We acknowledge support from the National Institute of Mental Health (Research Grants ROI MH42655, R24 MH51669; Research Scientist Development Awards KOl MH00849 and K02 MH012989 to the second author). We would like to thank Marisol Pena and Jose Calderon for their programming aeeietance, and Paul Cleary, Donald Steinwachs and Daniel Freeman for their advice and comments on various versions of this project.