To determine how coverage by Medicaid affects the use of preventive care, as well as the location and continuity of care for poor children. Analysis of the 1988 National Health Interview Survey on Child Health. Nationally representative sample of the US noninstitutionalized, civilian population. A total of 17,710 children under 18 years of age selected in a stratified cluster sampling of US households. None. The percentage of children with a usual source of routine care; the timeliness of visits for routine care; the usual source where routine care is received; and the continuity between sources of care. Poor children with Medicaid coverage were more likely than those without Medicaid to have a usual source of routine care (91% vs 78%, P less than .001) and to receive routine care within an appropriate time interval (84% vs 69%, P less than .001). However, poor children with Medicaid were less likely than children living above the poverty line to receive routine care in physicians' offices (56% vs 82%, P less than .001) and more likely to lack continuity between usual sources of routine and sick care (18% vs 6%, P less than .001). Children receiving routine care at community clinics compared with children receiving care at physicians' offices were more likely to receive sick care at a different location than where they receive routine care (40% vs 4%, P less than .001) and also more likely to identify an emergency department as their usual source of sick care (9% vs 2%, P less than .001). While Medicaid does improve access to care for poor children, it does not ensure them access to the same locations and continuity of care as that available to other children. Recent changes in the Medicaid program may address some of these inequities, but others are likely to remain.