The purpose of this study was to validate a version of the Working Alliance Inventory (WAI) for children and adolescents (WAI-CA). The sample included 109 children/adolescents aged between 7 and 17 years, outpatients in a Clinical Psychology Unit (Portugal), who completed the WAI-CA between psychotherapy sessions 3 and 35. A subsample of 30 children/adolescents aged between 10 and 14 years filled out both the WAI-CA and the WAI within a one-to-two week's interval. A subsample of 57 children/adolescents with ages between 7 and 17 years filled out the WAI-CA, and their accompanying parent the WAI. Results show high internal consistency (Cronbach's alpha ranging from.71 to.89) and good external validity. Significant differences were found in the bond subscale according to age, gender, and diagnosis, with higher values in children compared to adolescents, in girls compared to boys, and in participants with internalizing and externalizing problems compared to participants with school problems. Moderate to strong significant correlations were found between children/adolescents' WAI-CA and WAI scores and weak correlations between children/adolescents' WAI-CA scores and parent's WAI scores. Results suggest that the WAI-CA is a valid measure of working alliance to be used with children and adolescents.