Background: The World Health Organization estimates that 12 million children worldwide between the ages of 5 and 15 years have visual impairment owing to uncorrected refractive error. Aim: To assess whether the knowledge and practices of visual acuity (VA) screening improved after structured training of Grade 5 educators. Setting: Primary schools in Chatsworth, a suburb of Durban, South Africa. Method: A cluster randomised controlled trial was conducted. Bioethics approval and informed consent was obtained. Thirty-eight of 41 schools were randomised to an intervention or control group. Each Grade 5 classroom was given a Snellen chart, and educators from the intervention schools received structured training on VA screening and how to recognise visual impairment in learners. Data were collected from the 19 intervention and 18 control educators using a self-administered questionnaire at outset (baseline) and 6 weeks later in both groups. The Wilcoxon signed-rank test was used to measure the significance of changes. Results: The proportion of trained primary school educators who had adequate knowledge of VA screening increased significantly from 5.3% to 100% in the intervention group. In the control group, educators’ knowledge of VA screening stayed the same. The proportion of trained educators who performed VA screening of children increased from 0% to 79% (p < 0.001). Conclusion: A simple structured VA screening programme presented to primary school educators increased their knowledge and led to improved visual screening practices in the classroom.