Despite the wide use of the Parental Bonding Instrument, there is a controversy as to whether it should be viewed as measuring two or three factors. Recent studies using confirmatory factor analysis indicated that a three-factor model (warmth, protectiveness, and authoritarianism), proposed by Kendler (1996), provided the best fit to the data. The present study was designed to address the following question: Does the three-factor model provide a greater understanding of the relationships between the parental bonding factors and different types of depression when compared to the two-factor structure? Participants included 97 outpatients with early-onset dysthymic disorder (DD), 45 outpatients with episodic major depressive disorder (MDD), and 45 healthy control subjects. The results indicated that the three-factor model revealed associations between parenting behaviors and different types of depression that were not evident using the two-factor structure’s original dimensions.