Objective: The National Breast and Cervical Cancer Early Detection Program (NBCCEDP) provides free cancer screening to many low-income, underinsured women annually but does not routinely collect all data necessary for precise estimation of mammography rescreening rates among enrollees. Materials and Methods: To determine the percentages rescreened and to identify factors that encourage on-schedule rescreening, telephone interview and medical record data were collected from 1685 enrollees in Maryland, New York, Ohio, and Texas at least 30 months after their 1997 index mammogram. Results: Overall, 72.4% [95% confidence interval (95% CI) = 70.1-74.7] were rescreened within 18 months and 81.5% (95% CI = 79.6-83.5) within 30 months. At 30 months, the adjusted odds ratios (ORs) for rescreening were higher among Hispanics (OR = 1.95, 95% CI = 1.15-3.28), women with a history of breast cancer before the index mammogram (OR = 3.36, 95% CI = 1.07-10.53), and those who had used hormone replacement therapy before their index mammogram (OR = 1.94, 95% CI = 1.30-2.91). The 30-month adjusted ORs were lower for women who reported poor health status (OR = 0.60, 95% CI = 0.42-0.85), did not have a usual source of care (OR = 0.61, 95% CI = 0.40-0.94), did not know if they could have another free mammogram (OR = 0.28, 95% CI = 0.14-0.51), described their index screen as their first mammogram ever (OR for no prior mammograms versus three or more = 0.40, 95% CI = 0.27-0.60), did not recall receiving a rescreening reminder (OR = 0.35, 95% CI = 0.25-0.48), or did not think they had been encouraged to rescreen by their provider (OR = 0.61, 95% CI = 0.44-0.86). Discussion: Rescreening behavior in this sample of NBCCEDP enrollees was comparable with that observed in other populations. To facilitate routine rescreening among low-income women, ongoing efforts are needed to ensure they receive annual reminders and encouragements from their medical providers and that they know how to obtain the services they need.