SummaryThis paper reports an administrative follow‐up of 107 children at least 7 years old who were assessed on the Griffiths scales of mental development for clinical purposes when they were aged 4–4.11 years. When Griffiths scale scores are compared with educational status at follow‐up, the analyses show that groups of children in settings requiring different levels of learning capacity scored differently on the scales and in line with expectation. There is also evidence of this trend in earlier years. The clinical use of this trend is illust