Although high-quality early care and education (ECE) is widely accepted as one of the most effective means for promoting early learning and development, many ECE programs have limited impact perhaps because of issues with how ECE quality is defined and measured. This study seeks to expand definitions of ECE quality by asking which preschool ECE quality dimensions relate to gains in which developmental outcomes and contrasting measurements using classroom-level ratings and child-level behavior counts. The sample includes 366 children in 63 randomly selected prekindergarten (pre-k) classrooms in six rural counties in a Southeast state. In the fall and spring, children were administered tests of academic achievement, language skills, and executive functions, and teachers rated their social skills. In the winter, the quality of teacher-child interactions was rated at the classroom level and the frequency that the child experienced complex teacher talk, domain-specific instructional activities, and whole group settings were counted at the child level. Two sets of analyses related gains in outcomes to the classroom-level ratings in both analyses and to behavior counts summarized at the classroom level and at the child level in separate analyses. Results indicated that (1) different ECE dimensions related to gains in different outcomes; (2) ECE quality measures based on observing the selected experiences of individual children provide as strong or stronger associations with child outcomes than do ratings of teacher-child interactions; and (3) it may be necessary to measure experiences of individual children if those experiences are likely to vary markedly among children in the same classroom. (author abstract)