Does deferring school entry for children born just before the enrollment cutoff date improve their mental well-being? We address this question using administrative data on prescriptions for attention deficit hyperactivity disorder (ADHD) in England. Higher ADHD rates among early school starters are often attributed to a peer-comparison bias caused by differences in relative age among classmates. However, previous studies do not consider other potential underlying mechanisms. By adopting a more comprehensive framework, we can confirm that relative age is the primary driver of the gap in ADHD rate in the long term. Furthermore, we find that such a long-term gap is driven by first-time prescriptions between ages 5 and 8, which is a critical period when the accuracy of ADHD diagnosis is most important. Based on these findings, our policy recommendations include sorting children by age and refining diagnostic decision-making in early primary school.