The purpose of this study was to examine changes in students' relational engagement across the transition to high school in three schools reformed to improve the quality of student-teacher relationships. In order to analyze this data we employed latent growth curve (LGC) modeling techniques (n = 637). We ran three LGC models on three indicators of relational engagement: perceived teacher press (Middleton and Midgley in "Contemp Educat Psychol" 27:373-391, 2002), perceived teacher support (Goodenow in "Psychol Sch" 30:79-90, 1993a), and perceived school belonging (Goodenow in "J Early Adolesc" 13, 21-43, 1993b). For each model we estimated the latent means of the observed variables, the intercept, and the slope. Our hypothesized model for each of the variables included estimates for the contribution of team status (no team, cross-team, pure teamed), total absences in 8th grade, race (white/non-white), gender, gifted status, and socio-economic status. Alternate models were run dropping non-significant exogenous variables from the model one at a time. In each case, the hypothesized models, including five exogenous predictors, evidenced poor fit. However, alternate models including fewer exogenous predictors evidenced better fit, allowed for the estimation of intercept differences and rate of change, and accounted for significant amounts of variance in relational engagement. Findings are situated within the context of school reform designed to improve relational engagement.