Academic buoyancy is developed as a construct reflecting everyday academic resilience within a positive psychology context and is defined as students' ability to successfully deal with academic setbacks and challenges that are typical of the ordinary course of school life (e.g., poor grades, competing deadlines, exam pressure, difficult schoolwork). Data were collected from 598 students in Years 8 and 10 at five Australian high schools. Half-way through the school year and then again at the end of the year, students were asked to rate their academic buoyancy as well as a set of hypothesized predictors (self-efficacy, control, academic engagement, anxiety, teacher-student relationship) in the area of mathematics. Multilevel modeling found that the bulk of variance in academic buoyancy was explained at the student level. Confirmatory factor analysis and structural equation modeling showed that (a) Time 1 anxiety (negatively), self-efficacy, and academic engagement significantly predict Time 1 academic buoyancy; (b) Time 2 anxiety (negatively), self-efficacy, academic engagement, and teacher-student relationships explain variance in Time 2 academic buoyancy over and above that explained by academic buoyancy at Time 1; and (c) of the significant predictors, anxiety explains the bulk of variance in academic buoyancy.