The aims of this study were (1) to investigate whether 6-7-year-old children are accurate in perceiving their actual movement competence, and (2) to examine possible age- and gender-related differences. A total of 603 children (301 girls and 302 boys, aged 6 to 7 years) were assessed on the execution accuracy of six locomotor skills and six object control skills using the Test of Gross Motor Development (TGMD-2). The perceived competence of the same skills, plus six active play activities, was also gauged through the Pictorial Scale of Perceived Movement Skill Competence (PMSC-2). The factorial validity of the TGMD-2 and PMSC-2 scales was preliminarily ascertained using a Bayesian structural equation modeling approach. The relationships between the latent factors of the two instruments were then assessed. Gender and age differences were also examined. The factorial validity of the TGMD-2 and the PMSC-2 was confirmed after some adjustments. A subsequent analysis of the relationship between the latent factors (i.e., locomotor skills and object control) of the two instruments yielded very low estimates. Finally, boys and older children showed better competence in object control skills compared to their counterparts. Weak associations between actual and perceived competence suggest that inaccuracy in children's perceptions can be likely due to a still limited development of cognitive skills needed for the evaluation of the own competence. From an applied perspective, interventions aimed at improving actual motor competence may also increase children's self-perceived motor competence and their motivation toward physical activity.