Background: In spite of the importance of maternal sensitivity as a construct that fosters secure attachment and promotes a child’s social and cognitive development, no routine clinical screening currently identifies mothers at risk of poor maternal sensitivity. This is partly because researchers have not identified all the factors that influence maternal sensitivity. As a result, parenting interventions to promote maternal sensitivity and optimise child outcomes tend to focus on clinical groups. Thus, more attention is needed to identify possible determinant factors. The neurobiological mechanisms underlying natural variation in maternal sensitivity (i.e. sensitive and less sensitive mothers) are poorly understood, especially the putative role of the hormone Oxytocin (OT). Literature has suggested that this variation in maternal sensitivity is an outcome of interaction between maternal OT, as well as social factors (e.g. perceived parenting) and this interaction charts the discrete profile of the maternal brain that is mediated by stress- and reward-related neural systems. To date no study examined for the neurobiological correlates of maternal sensitivity in a distinct group of mothers representing natural variations in maternal sensitivity. Methods: Out of 105 women recruited from community antenatal clinics during their pregnancy, to complete a set of self-reported questionnaires assessing their psychosocial characteristics, a total of 80 new (i.e. early postpartum) mothers and their infants were followed up and underwent evaluation of maternal sensitivity at 4-6 months postpartum. Using a stepwise regression, we examined for predictors of maternal sensitivity among the sample (Study I). Later, at 7-9 months postpartum, 30 mothers, representing extremes in maternal sensitivity, were selected from this sample of 80: 15 mothers with higher scores (high sensitivity mothers - HSMs), and 15 with lower scores for maternal sensitivity (low sensitivity mothers - LSMs), underwent functional Magnetic Resonance Imaging (fMRI) to examine their brain responses when viewing videos of their own and an unknown infant. Maternal plasma OT levels were also measured before and following an interactive play with their infant (Study II). Results: Mothers’ self-reported experience of own parental care, and household income, independently predicted maternal sensitivity, accounting for 17% of the variance. Comparing mothers grouped by maternal sensitivity level, HSMs showed a drop in their plasma OT levels following the interaction with their infant. HSMs also showed significant brain activation in the right superior temporal gyrus in response to own infant (compared to unknown infant) when compared to LSMs. By contrast LSMs did not show any change in their plasma OT levels following interaction with their infant, and their brain responses to own infant did not show any significant brain activation when compared to HSMs. Conclusions: The findings may have implications for future novel approaches for early assessment of mothers at risk of low maternal sensitivity so they could be targeted by specialised assessments and consequently interventions to improve their parenting (Study I). Maternal sensitivity is accompanied by neural correlates that could act as a biomarker for future intervention studies that target vulnerable mothers (Study II).