Child care stability affects child and family outcomes. Stability reflects the time dimension of a child care arrangement. Although stability does not guarantee positive outcomes, instability appears to decrease the likelihood of achieving them. Some level of stability is a necessary, although not sufficient, characteristic of care that meets children's needs. Child care stability is of special concern for children in low income families because child care impacts are greatest for these children, and current welfare policies result in more low-income children in nonparental care. This study increases understanding of child care stability through (a) an analysis of findings from stability studies over 30 years, (b) an examination of relationships of the four major stability measures, and (c) presentation of results from an analysis of the stability of subsidized child care arrangements in Oregon. The analysis of stability studies documented lack of consistency in conceptualization, measures, and methodology. These inconsistencies limit comparisons of reported stability findings and confidence in estimates of child care stability that have been reported. Examination of the four stability measures found that the three child-level measures appear to describe the same construct as they are highly correlated. The fourth stability measure is at the level of the arrangement and captures a distinctly different aspect of stability. Stability levels of subsidized arrangements in Oregon appear lower than those found in nationally representative samples but similar to levels found in populations participating in public assistance programs. About a third of children had very stable care but others had high levels of instability. Fifty percent of arrangement spells ended by 3 months, even when children were observed for 36 months. The study concludes with recommendations for future research. / Graduation date: 2005