Pregnant women with psychosocial vulnerabilities should be offered perinatal interventions that include a parenting component to ameliorate the potential negative effects of maternal mental health problems and/or poor social network on parenting. One such intervention program is the Circle of Security-Parenting intervention (COS-P). The COS-P is a manualized video-based intervention that based on attachment theory seek to enhance maternal sensitivity and decrease the risk on insecure and disorganized attachment. We carried out a randomized controlled trial examining the efficacy of a perinatal adapted version of COS-P for women with psychosocial vulnerabilities (e.g. histories of mental health problems and/or poor social networks). Eligible participants (N = 78) were recruited to the study by midwives during regular prenatal sessions. Interventions were delivered individually at home by trained health nurses both pre and post birth. The primary outcome was maternal sensitivity assessed with the Coding Interactive Behavior Manual by blinded coders from video-recordings of mother-infant free play interactions. Secondary outcomes were mother-reported depressive symptoms, parental reflective functioning, parental stress, infant socio-emotional functioning, and maternal wellbeing. All outcomes were assessed at nine months infant age. We did not find an effect of the intervention on the primary outcome of maternal sensitivity (β = -0.08; 95% CI [-0.41, 0.26], p = .66). Neither did we find intervention effects on the secondary outcomes of depressive symptoms, parental reflective functioning, maternal well-being, or infant socio-emotional functioning. We did however find that the intervention decreased parental stress (β = -8.51; 95% CI [-16.6;-0.41], p = .04). The results are discussed in light of existing findings on the effect of COS-P and sample heterogeneity. Furthermore, we discuss the challenges of adapting the COS-P for pregnant women, some without prior experiences with caregiving. Future research with larger at-risk samples examining moderation factors (e.g. adult attachment, depression maternal-fetal attachment) are recommended.