Most US students attend coeducational classes, but to what extent do students feel integrated into the entire classroom of their peers, especially with other-gender (OG) peers? The major goal of this study was to investigate how variations in gender integration (GI), measured by students' expectancies about inclusion, efficacy, and social costs of interacting with OG peers, predicted school liking and classroom supportiveness over an academic year, using a short-term longitudinal design. We also explored how students' expectancies changed over the year. Participants included elementary school students (515 school-age children; 51% boys, Mage = 9.08 years, SD = 1.00; 3-5th grade; 26 classrooms). A two-wave latent change score model showed that changes over the year varied depending on the type of expectancy, grade, and gender, with decreases in inclusion and efficacy for boys. Longitudinal path analyses conducted to assess whether GI expectancies predicted school belongingness showed that students' levels of OG inclusion in the Fall uniquely predicted changes in levels of school liking and classroom community over the year, even with many controls in the model. The findings demonstrate that students' relationships with OG peers matter for having a sense of belonging in school, and educators should support and encourage these relationships. [For the Grantee submission, see ED618764.]