First-generation (1st-gen) college students face unique obstacles that threaten to erode their academic motivation and success during the school-to-college transition (Stebleton & Soria, 2012). Attribution-based motivation treatments can improve achievement for failure-prone college students (Perry & Hamm, 2017), yet their efficacy for students with socioeconomic academic risk factors remains unexamined. The present longitudinal, pre-post, randomized treatment field study administered attributional retraining (AR) to 1st-gen and second-generation (2nd-gen) college students in an online two-semester introductory course who differed in academic control beliefs (low, high). 1st-gen, low control AR recipients outperformed their no-AR peers by a full letter grade (B vs. C+), and were 48% less likely to drop out of the course. Conditional process analyses revealed that AR-achievement linkages were mediated by causal attributions and perceived control in a hypothesized causal sequence. Results further the literature by demonstrating that AR can boost the achievement of at-risk 1st-gen students indirectly via motivation-related variables specified within Weiner’s (1985, 2014, 2018) attribution theory.