Given the prevalence of short child care subsidy spells and program churning documented in prior studies, researchers and policymakers have been concerned about the implications of discontinuity in subsidy receipt for the stability of children's care arrangements. Yet little research has studied the stability of subsidized arrangements or how subsidy discontinuity relates to changes in subsidized providers. Using child care subsidy program administrative records from a cohort of children in four diverse sites across Illinois and New York states, this study examines patterns of subsidy use and stability of subsidized care arrangements, as well as the relationship between the two. Results suggest that the length of states' eligibility periods is related to the duration of subsidy spells; however, significant variation in patterns of subsidy use within states suggests that local level factors are also important. Results show that subsidy discontinuity is related to children experiencing more total changes in subsidized providers. Focusing on provider changes across spells, we also find that the timing of subsidy exits, the length of gaps in subsidy receipt, and within spell provider instability are each related to whether or not children re-enter the program with a different subsidized provider after a break in subsidy receipt. We discuss these findings' implications for understanding how new program requirements established in the 2014 reauthorization of the Child Care and Development Block Grant may matter for subsidy continuity and care stability. (author abstract)