Mothers with prenatal substance use disorder (SUD) often show broad deficits in their reflective functioning (RF), implying severe risk for the relationship with their baby. Two different types of prenatal maternal RF may be important for parenting: adult attachment-focused-RF (AAI-RF), regarding parent's own childhood experiences, and parenting-focused RF (PRF) regarding their own current process of becoming a parent. However, their inter-relations and potentially different roles for parenting intervention outcomes are not clear. This study examined the associations between mothers' prenatal AAI-RF and pre- and post-natal PRF, and their role in mother-infant interaction and substance use as treatment outcomes. The participants were 57 treatment-enrolled pregnant mothers with SUD and 50 low-risk comparison mothers. AAI-RF was measured with the Adult Attachment Interview. For a subsample of 30 mothers with SUD, PRF was measured with Pregnancy Interview (during pregnancy/pre-intervention), and with Parent Development Interview at 4 months (during intervention). Mother-infant interaction was measured with Emotional Availability Scales at 4 and 12 months (post-intervention), and maternal substance use by post-natal substance relapses. Prenatal AAI-RF and pre- and post-natal PRF were highly associated with each other. Only higher prenatal PRF predicted better mother-infant interaction quality at 4 months and less substance use during the child's first year. Interestingly, prenatal PRF and AAI-RF predicted opposite changes in mother-infant interaction: lower prenatal PRF, but higher AAI-RF predicting more positive change. AAI-RF was especially associated with a change in maternal intrusiveness and hostility, indicating that it represents a more general regulatory tendency. Further studies are needed in larger and lower-risk samples. Our results suggest, however, that AAI-RF and PRF are partially distinct and should be uniquely targeted in perinatal interventions.