The findings of the study indicate that memories of childhood boundary disturbances can predict boundary disturbances across generations. Two main types of boundary disturbance patterns were observed from the family interactions that were assessed when children were 24 months old: (1) enmeshed interactions, in which parents employed guilt-inducing or coy behavior to induce children to meet his or her own needs, and (2) controlling interactions, in which parents used power-assertive tactics to gain compliance. Enmeshed boundary disturbances were predicted by two main variables: (1) by fathers’ memories of role-reversed relationships with their mothers, and (2) by fathers’ hostile and role-reversed infant caregiving patterns. Controlling boundary disturbances were predicted mainly by mothers’ current mental representations of attachment. Research methodology involved three phases. In the first phase, 96 parents who were expecting their first child were interviewed about their child-parent relationships with their own parents as well as their current representations of attachment relationships. During phase 2 of the research, mothers and fathers were videotaped separately interacting with their 8-month-old infants in their home while phase 3 of the research involved the families participating together in a series of triadic interaction tasks with their now 24-month-old toddlers. The findings from this study should be considered speculative and must be examined using larger samples as well as clinical samples. Tables, references