The health care system for the poor is complex and fragmented, especially with regard to pregnancy and other reproductive health services. To obtain reproductive health care and other primary health care services, poor women may have to interact with a variety of facilities and practitioners. It seems likely that this fragmentation would discourage utilization of preventive care, compromise the quality of care that is obtained, and result in increased costs when services are delayed or repeated because of the lack of coordinated care. The purpose of this study is to describe the utilization of different types of health care by pregnant Medicaid recipients and examine the relationship of different patterns of utilization to the amount, costs and quality of maternity care. The population under study will be a sample of continuously enrolled Maryland Medicaid recipients who gave birth in 1988, and the sources of data will include Medicaid claims, birth certificates, and medical chart abstracts. The study will describe utilization in terms of its longitudinality, defined as the existence of a long-term relationship with a provider spanning illness episodes and problems, and continuity, the degree to which continuous care is given by providers or practitioners within a specific episode of illness. This study has the potential to add to research evidence about the effects of continuity and longitudinality on the quality and costs of maternity care and to suggest ways to improve the structure of health care delivery for the poor.