To ensure that researchers and practitioners make valid interpretations from children's self-reports, evidence must demonstrate that the self-report has appropriate psychometric properties. The Child Occupational Self Assessment (COSA) is a self-report of occupational competence and value for everyday activities designed to involve children in identifying goals and assessing outcomes. Five hundred two children with disabilities, ages 6-17, completed the COSA. We used a Rasch Partial Credit model and parametric and nonparametric statistics to obtain validity evidence. Evidence suggests that the COSA has good content, structural, and substantive validity as given by item and child fit statistics and unidimensionality evaluation. Evidence for external validity was mixed because child fit status and measures varied with some demographic and assessment administration variables. Evidence suggests that most children's responses to the COSA can be validly interpreted as indicators of occupational competence and value for everyday activities.