Research on individual differences demonstrates that children's perceived control exerts a strong effect on their academic achievement and that, in turn, children's actual school performance influences their sense of control. At the same time, developmental research shows systematic age-graded changes in the processes that children use to regulate and interpret control experiences. Drawing on both these perspectives, the current study examines (1) age differences in the operation of beliefs-performance cycles and (2) the effects of these cycles on the development of children's perceived control and classroom engagement from the third to the seventh grade. Longitudinal data on about 1,600 children were collected six times (every fall and spring) over 3 consecutive school years, including children's reports of their perceived control and individual interactions with teachers; teachers' reports of each student's engagement in class; and, for a subset of students, grades and achievement tests. Analyses of individual differences and individual growth curves (estimated using hierarchical linear modeling procedures) were consistent, not only with a cyclic model of context, self, action, and outcomes, but also with predictors of individual development over 5 years from grade 3 to grade 7. Children who experienced teachers as warm and contingent were more likely to develop optimal profiles of control; these beliefs supported more active engagement in the classroom, resulting in better academic performance; success in turn predicted the maintenance of optimistic beliefs about the effectiveness of effort. In contrast, children who experienced teachers as unsupportive were more likely to develop beliefs that emphasized external causes; these profiles of control predicted escalating classroom disaffection and lower scholastic achievement; in turn, these poor performances led children to increasingly doubt their own capacities and to believe even more strongly in the power of luck and unknown causes. Systematic age differences in analyses suggested that the aspects of control around which these cycles are organized change with development. The beliefs that regulated engagement shifted from effort to ability and from beliefs about the causes of school performance (strategy beliefs) to beliefs about the self's capacities. The feedback loop from individual performance to subsequent perceived control also became more pronounced and more focused on ability. These relatively linear developmental changes may have contributed to an abrupt decline in children's classroom engagement as they negotiated the transition to middle school and experienced losses in teacher support. Implications are discussed for future study of individual differences and development, especially the role of changing school contexts, mechanisms of influence, and developmentally appropriate interventions to optimize children's perceived control and engagement.