We use the Early Childhood Longitudinal Study to investigate national patterns addressing (a) who repeats kindergarten, and (b) the subsequent cognitive effects of this event. Using OLS regression techniques, we investigate 1st-time kindergartners who are promoted, 1st-time kindergartners who are retained, and children who are already repeating kindergarten. Boys, children from low socioeconomic backgrounds, and children who enter kindergarten younger than typical age standards are consistently at risk for repeating kindergarten, but racial/ethnic patterns differ across the 2 cohorts of kindergarten repeaters. Evidence suggests that repeating kindergarten rarely leads to cognitive benefits in literacy or mathematics performance. On average, kindergarten repeaters continue to perform below their peers in terms of literacy skills both at the end of kindergarten and at the end of first grade (effect size [ES] = -0.20 and -0.24, respectively). In mathematics, the performance differentials are smaller but remain statistically significant. Evidence suggests that these differences vary somewhat by children's background and the school setting. Most children appear to receive little or no cognitive benefit from repeating kindergarten, suggesting the need for a careful reconsideration of current retention practices. (Contains 9 tables and 4 footnotes.)