There is variability in home visiting program impacts on the outcomes achieved by high risk families. An understanding of how effects vary among families is important for refining service targeting and content. The current study assessed whether and how maternal attributes, including relationship security, moderate short- and long-term home visiting impacts on maternal psychosocial functioning. In this multisite RCT of home visiting in a population-based, ethnically-diverse sample of families at risk for maltreatment of their newborns (n = 643), families were randomly assigned to home visited (HV) and control groups. HV families were to receive intensive services by trained paraprofessionals from birth-3 years. Outcome data were collected when children were 1, 2, and 3 years old and 7, 8, and 9 years old. Overall, short- and long-term outcomes for HV and control mothers did not differ significantly. Demographic attributes, a general measure of overall maternal risk, and partner violence did not moderate program impact on psychosocial functioning outcomes. Maternal relationship security did moderate program impact. Mothers who scored high on relationship anxiety but not on relationship avoidance showed the greatest benefits, particularly at the long-term follow-up. Mothers scoring high for both relationship anxiety and avoidance experienced some adverse consequences of home visiting. Further research is needed to determine mediating pathways and to inform and test ways to improve the targeting of home visiting and the tailoring of home visit service models to extend positive home visiting impacts to targeted families not benefiting from current models.