This article describes the development and psychometric testing of the Connection-Disconnection Scale (CDS), the only self-report measure designed to assess perceived mutuality experienced in close relationships by women with eating disorders. Item development was informed by relational-cultural theory and focus groups with patients and recovered individuals. Content and construct validity, test–retest reliability, and internal consistency were examined in samples of female outpatient (n = 131) and partial hospitalization (n = 85) patients. Factor analysis with promax rotation for each version of the CDS in outpatients resulted in a single-factor model explaining 77.53% of the variance for CDS-Mother, 71.86% for CDS-Father, 77.79% for CDS-Partner, and 67.67% for CDS-Friend. Cronbach’s alphas ranged from .97 to .99 for both samples. Overall, the CDS demonstrated good discriminant and convergent validity with moderate to strong correlations between CDS parent forms and the Parental Attachment Questionnaire, Dyadic Adjustment Scale, and the Social Support Questionnaire. Regression equations revealed that scores on CDS parent forms were inversely related to several subscale scores on the Eating Disorders Inventory—2. The CDS is a reliable and valid measure of perceived mutuality that can enrich relational understanding of the etiology and treatment of eating disorders in women.