SUMMARY STATEMENT Parental care is an innate social behavior essential for species survival. Hence, a hardwired circuit has evolved to support its expression. However, naïve animals often show hostile behaviors toward infants, including infanticide. These are thought to be selected to free up resources for the perpetrator's own future offspring. Though infanticide was once considered pathological, its prevalence—e.g., in nearly 100% of wild mice—suggests that it is a part of animals' natural behavior repertory and could be supported by a dedicated neural circuit. We reason that hostile behaviors toward the young under pathological conditions could be caused by deficits in both parental and infanticidal circuits. Thus, a better understanding of both circuits is needed to ultimately prevent child abuse. Our earlier study pinpointed estrogen receptor alpha (Esr1)- expressing cells in the medial preoptic area (MPOAEsr1) as an essential population for maternal care. More recently, we demonstrated Esr1 cells in the principal nucleus of the bed nucleus of stria terminalis (BNSTprEsr1) as a critical population driving infanticide in female mice. MPOAEsr1 and BNSTprEsr1 cells antagonize each other and change excitability in opposite directions during motherhood to support the emergence of maternal behaviors. Here, building on these findings, we will continue our efforts to elucidate the neural circuits underlying infant-directed hostile and caring behaviors using BNSTprEsr1 and MPOAEsr1 cells as two anchoring points. In Aim 1, we will investigate the role of BNSTprEsr1 cells in infanticide in male mice. This is an important question given that BNSTpr is sexually dimorphic anatomically and molecularly. We will also examine the relationship between male BNSTprEsr1 cells activated during infant-directed behaviors and those activated during adult-directed social behaviors using two-photon calcium imaging. In Aim 2, we will further dissect the infanticide circuit downstream of BNSTprEsr1 cells using pathway-specific manipulation tools. In Aim 3, we will investigate the role of medial amygdala (MeA), an upstream region of both MPOA and BNSTpr, in young-directed behaviors. We hypothesize that MeA cells modulate positive and negative infant-directed behaviors through their projections to the BNSTpr and MPOA, respectively. Overall, this study promises to provide new information about the neural underpinning of infant-directed behaviors. Such knowledge is an important step toward new strategies to prevent child abuse in humans that can profoundly impair child health and development.