Classroom activities aimed at changing students' identity-based motivation (IBM) improve student outcomes by helping students experience school as the path to their adult future identities and their difficulties along the way as signals of the importance of schoolwork. One way to scale these effects would be to have teachers deliver IBM activities. Hence, we asked if, after brief two-day training, teacher-delivered IBM intervention could meet fidelity standards and if attaining more fidelity matters. We trained all eighth-grade teachers in two middle schools (N?=?211 students). We used Dane and Schneider's (1988) five-component fidelity model and Durlak and DuPre's (2008) empirically derived threshold and practical maximum standards for fidelity. We found that most classrooms (88%) and students (89%) received IBM intervention at-or-above threshold standard, implying that teacher-based IBM delivery is viable. Moreover, investing in improving fidelity is worthwhile; above-threshold fidelity improved core grade-point-average and reduced risk of course failure.