BACKGROUND: Diagnostic magnetic resonance imaging (MRI) criteria have not been sufficiently validated in pediatric multiple sclerosis (MS) despite differences in epidemiologic data and clinical disease courses between pediatric and adult MS. OBJECTIVE: The objective of this paper is to evaluate the diagnostic applicability and validity of the revised McDonald diagnostic criteria 2010 in a large cohort of pediatric-onset MS patients (POMS) and compare them to previously recommended MRI-based classifications. Furthermore, we aimed to investigate the contribution of spinal cord lesions to the revised McDonald criteria 2010. METHODS: A cohort of 85 patients with definite MS, age at onset 2.8-18 years, was analyzed in a retrospective multicenter study. Number and regional distribution of T2w and contrast-enhancing T1w lesions at initial and follow-up MRIs were main outcome measures. RESULTS: In 62% of POMS the initial MRI within four weeks after symptom onset was sufficient to diagnose MS according to the revised McDonald criteria 2010. In a subcohort of patients with spinal MRI at first presentation, 10% reached the dissemination in space (DIS) and dissemination in time (DIT) criteria only by the inclusion of contrast-enhancing spinal lesions. CONCLUSIONS: The revised McDonald criteria 2010 facilitate the diagnosis of POMS already at first presentation. The addition of a spinal cord MRI was helpful only in selected cases.