The primary retention of molars observed in clinic corresponds to a still-unexplained 29 absence of molar eruption despite the presence of an eruption pathway, resembling the 30 experimental transient inhibition of RANKL signaling in mice. The aim of the present study was to 31 confront the hypothesis according to which the primary retention of molars is associated with 32 transitory perturbations to RANKL signaling during growth as part of a wider craniofacial skeleton 33 pattern. The experimental strategy was based on combining a clinical study and an animal study 34 corresponding to characterization of the craniofacial phenotypes of patients with primary retention 35 of molars and analyses in mice of the consequences of transient inhibition of RANKL signaling on 36 molar eruption and craniofacial growth. The clinical study validated the existence of a particular 37 craniofacial phenotype in patients with primary retention of molars: a retromandibular skeletal 38 class II typology with reduced mandibular dimensions, which manifests itself at the dental level by 39 a class II/2 with palatoversion of the upper incisors and anterior overbite. The animal study 40 demonstrated that transient invalidation of RANKL signaling had an impact on the molar eruption 41 process, the severity of which was dependent on the period of inhibition, and was associated with a 42 reduction in two craniofacial morphometric parameters: total skull length and craniofacial vault 43 length. In conclusion, primary retention of molars may be proposed as part of the craniofacial 44 skeleton phenotype associated with a transitory alteration in RANKL signaling during growth. 45