Background: A consensus of early treatment with disease-modifying therapies (DMT) in multiple sclerosis (MS) has been reached based on several observational and experimental studies in adults. However, paediatric onset (PO)MS appears phenotypically different from adult onset MS, characterized by increased relapse rate and pronounced radiological activity on MRI. The objective of this study was to investigate the long-term consequences of delayed treatment start in POMS on disability in a real-world, population-based setting. Methods: Based on prospectively collected data from The Danish Multiple Sclerosis Registry, we defined a cohort of MS patients with onset before the age of 18 years, who were born in 1980 or later, and started treatment with a DMT between 1998 and 2018. The POMS cohort was stratified according to treatment start within 2 years of onset (N = 140) or later (N = 151). Annualised relapse rate in each study group was compared using a negative binomial regression; and Cox proportional hazard model was used to estimate hazard ratios (HR) of time to sustained Expanded Disability Status Scale (EDSS) score 4, 6-month confirmed EDSS worsening and 6-month confirmed EDSS improvement, respectively, according to disease duration. Results: The POMS cohort had a total median follow-up time of 7.7 years (interquartile range 4.6–11.6). There was no association between risk of relapses in patients with delayed treatment start compared to earlier treatment start. Patients starting on a DMT later than 2 years after onset had a 2.52-fold increased risk of reaching sustained EDSS 4 compared to those starting within 2 years of onset (HR=2.52, 95% confidence interval (CI)=1.01–6.34). For every year increment from onset to start of first DMT, the risk of reaching sustained EDSS 4 increased by 17% (HR=1.17, 95% CI=1.05–1.30). In line with this, the risk of reaching confirmed EDSS worsening was increased by 44% compared to those starting earlier, although not statistically significant (HR=1.44, 95% CI=0.95–2.19). Starting on a DMT later was associated with 61% decreased chance of confirmed EDSS improvement compared to those starting earlier (HR=0.39, 95% CI=0.26–0.59). For every year increment from onset to starting DMT, the risk of confirmed EDSS improvement decreased by 10% (HR=0.90, 95% CI=0.84–0.96). Conclusions: Delayed treatment start in this POMS cohort was associated with shorter time to reach sustained EDSS 4 and confirmed EDSS worsening, and decreased chance of reaching confirmed EDSS improvement, and thus support early treatment start in POMS patients.