This study explored the relative salience of students' personal strengths and teacher support in predicting academic risk status. The participants were Secondary One (S1, Grade Seven) students from Singapore who scored below the cohort's mean score in a national test administered at the end of primary education, and were identified as low risk (n = 309) or high risk (n = 396), based on their S1 achievement score in the English Language subject. Logistic regression analysis was conducted with academic risk status as criterion variable and the following potential predictors: students' background variables (i.e. socio-economic status, cognitive ability and initial achievement), personal strengths, teacher-student relatedness, and teacher autonomy and competence support. After controlling for the effects of the students' background variables, teacher trust emerged as the strongest (negative) and most stable predictor of high-risk status; teacher alienation and teacher-student communication were found as significant positive predictors of students' placement in the high-risk group.