Early interventions aim at promoting a good mother-child relationship as basis for a good socio-emotional development, especially in high-risk populations, and at correcting already unfavorable patterns of interaction and are common today. Insecure attachment, both of the child and of the mother, has been identified as a risk factor for early regulation disorders and further child development. Based on accumulating evidence of effectiveness of animal-assisted interventions, we implemented an Equine-Assisted Intervention (EAI) as an early intervention approach for mother-child-dyads and investigated its effects in comparison to a conventional play-based early intervention (PBI) in a randomized controlled trial. Both interventions took place once a week for 45 minutes over a period of 8 weeks and aimed at improving maternal caregiving and security in the mother-child relationship. Twenty mother-child dyads (children aged 12 to 24 months) from a high-risk background, in some of which the child already displayed dysregulation symptoms, were investigated. Attachment of mother and child, maternal caregiving, and interaction within the dyad were assessed before and after the intervention. There was not statistically significant difference (p < 0.05) between intervention groups with regard to changes from insecure to secure attachment or from disorganized to organized attachment, assessed via the Ainsworth Strange Situation Test. The only significant group-difference with regard to self-reported caregiving, was found for proximity maintenance in mothers without previous therapy experience (n = 8), those in the EAI reaching higher scores (p = 0.064). Play interaction, assessed via the CARE-Index, showed that mothers and children in PBI improved significantly more in their interaction behavior on several scales, while child's difficultness only significantly improved in EAI.