The psychometric properties of the Autonomy and Relatedness Inventory (ARI)--a measure of the quality of primary intimate relationships--was assessed. This 30-item instrument was designed to assess both positive and negative dimensions of any type of dyadic relationship. Cross-sectional data from in-home interviews with 213 mothers with 5- and 6-year-old children were analyzed. Data derived from a separate scale on the quality of marital relationships for a subsample of 40 married mothers also were used to assess the ARI's validity. The ARI demonstrated strong internal consistency. Maternal characteristics of age, education, income, and number of children were not correlated with ARI scores. There also were no differences in ARI scores by marital status, race, or employment status of the mother. A principal components analysis with varimax rotation reduced the items to two orthogonal dimensions: (1) support/positive regard; and (2) dominance/control. Both dimensions of the ARI were discriminable from measures of depressive and psychosomatic symptoms of the mothers. Convergent validity was supported by significant correlations between ARI subscales and subscales of Spanier's Dyadic Adjustment Scale for a subsample of 40 married mothers. This preliminary evidence suggests that the ARI holds potential for use in future studies as a measure of the quality of primary intimate relationships. A 33-item list of references and five data tables are included. (Author/SLD)