Objectives: Modified adult disease severity scoring systems are being used for childhood FMF. We aim to test the clinical consistency of two common severity scoring systems and to evaluate the correlation of scores with the type of FMF mutations in paediatric FMF patients since certain mutations are prone to severe disease.Methods. Two hundred and fifty-eight children with FMF were cross-sectionally studied. Assessment of the disease severity was performed by using the modified scoring systems of Mor et al. and Pras et al. Genetic analysis was performed using PCR and restriction endonuclease digestion methods for the presence of 15 FMF gene mutations. FMF mutations were grouped into three based on well-known genotypic-phenotypic associations. Correlation between the mutation groups and the severity scoring systems was assessed. The consistency of the severity scoring systems was evaluated.Results. The results of two scoring systems were not statistically consistent with each other (κ = 0.171). This inconsistency persisted even in a more homogeneous subgroup of patients with only homozygote mutations of M694V, M680I and M694I (κ = 0.125). There was no correlation between the mutation groups and either of the scoring systems (P = 0.002, r = 0,196 for scoring systems of Mor et al.; P = 0.009, r = 0.162 for Pras et al.).Conclusions. The inconsistency of the two scoring systems and lack of correlation between the scoring systems and mutation groups raises concerns about the reliability of these scoring systems in children. There is a need to develop a scoring system in children based on a prospective registry. © The Author 2011. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved.