This study used a person-centered approach to identify school readiness profiles in a sample of kindergartners (n=1,826) from a large and diverse school district in the United States. Using latent profile analyses and multi-level modeling, we examined three aims: 1) whether patterns of readiness skills at kindergarten entry could be detected, 2) the extent to which detected patterns predicted gains in academic and social-emotional skills, and 3) whether the quality of teacher–child interactions moderated the associations between profile membership and end of kindergarten outcomes. Based on a comparison of fit indices, a 4-profile solution best represented the data. Eighteen percent of children were in the “High Risk” profile, 34 % were in the “Ready” profile, 20 % percent of children were in the “Social-Emotional Risk” profile, and 28 % of children were in the “High Readiness” profile. For all outcomes, we found that profile membership predicted spring scores, after controlling for fall scores of each skill, suggesting that the constellation of kindergarten readiness skills matters more than any one skill. We also found that the quality of teacher–child interactions moderated associations between profile membership and changes in achievement and problem behavior for children within some groups. Results suggest that children enter kindergarten with unique profiles of skills and that supports for children may need to be individualized based on profiles to maximize social, emotional, and academic development. Results also suggest that high-quality teacher–child interactions may serve as an important protective and promotive factor for children with certain clusters of readiness skills at kindergarten entry. (author abstract)