Universities in South Africa face ongoing challenges with low rates of academic attainment and high rates of attrition. Our aims were to (1) investigate the extent to which common mental disorders evaluated early in the first year predict academic failure at the end of the year, controlling for sociodemographic factors and (2) establish the potential reduction in prevalence of failure that could be achieved by effectively treating associated mental disorders. Self-report data were collected from first-year students (n = 1402) via an online survey at the end of the first semester. Participants were assessed for six common mental disorders. Academic performance data were subsequently obtained from institutional records at the end of the year. Bivariate and multivariate logistic regression models were used to identify the best sociodemographic and mental health predictors of academic failure. Population attributable risk analysis was used to assess the potential impact of treating associated mental disorders. In multivariate logistic regression models controlling for significant sociodemographic factors, the odds of failure were elevated among students with major depressive disorder (aOR = 3.69) and attention deficit hyperactivity disorder (aOR = 2.05). Population attributable risk analysis suggests that providing effective treatment to students with major depressive disorder and/or attention deficit hyperactivity disorder could yield a 6.5% absolute reduction in prevalence of academic failure (equivalent to a 23.0% proportional reduction in prevalence of academic failure). Providing effective and accessible campus-based mental healthcare services is integral to supporting students’ academic attainment and promoting transformation at South African universities.

南非的大学面临着持续的挑战，即学历率低和损耗率高。我们的目标是（1）调查第一年初期评估的常见精神障碍在多大程度上预测了年底的学业失败，控制社会人口统计学因素，以及（2）确定潜在的患病率降低可能是通过有效治疗相关的精神障碍实现。自我报告数据来自一年级学生（n = 1402），这是通过第一学期末的在线调查得出的。对参与者进行了六种常见精神障碍的评估。随后在年底从机构记录中获得了学业成绩数据。使用双变量和多元逻辑回归模型来确定最佳的社会人口统计学和学术研究的心理健康预测指标。人群归因风险分析用于评估治疗相关精神障碍的潜在影响。在控制重要社会人口统计学因素的多元逻辑回归模型中，患有重度抑郁症（aOR = 3.69）和注意缺陷多动障碍（aOR = 2.05）的学生的失败几率增加。人群归因风险分析表明，为重度抑郁症和/或注意缺陷多动障碍的学生提供有效的治疗可以使学业失败的患病率绝对降低6.5％（相当于学业失败的患病率成比例降低23.0％）。提供有效且可访问的校园心理医疗服务是支持学生的学术成就并促进南非大学转型的必不可少的部分。