Background Animal-assisted interventions and robotic animal interventions are becoming increasingly popular to support the care of people with dementia and may have the potential to improve a range of psychosocial outcomes. This review aims to identify, describe, and compare animal-assisted and robotic animal interventions delivered to people with dementia, their characteristics, effectiveness, and the proposed mechanisms underlying any potential impact. Methods A systematic literature search was conducted in MEDLINE, AMED, EMBASE, PsycINFO, OVID Nursing, PubMed, CINAHL and Web of Science. Random-effects meta-analyses of randomised controlled trials (RCTs) were conducted to summarise studies that evaluated common outcomes (agitation, depression, quality of life). A narrative approach was used to synthesise other findings. Results Fifty-one studies were included: 18 RCTs; 12 non-randomised trials, 13 cohort studies, 7 qualitative studies and one mixed-methods study. Meta-analyses were conducted for a small number of RCTs, with effectiveness of animal-assisted interventions demonstrated for agitation. Narrative findings suggested animal-assisted and robotic animal interventions may be promising in improving depression, agitation, and quality of life. Three potential mechanisms of action were identified for both animal-assisted and robotic animal interventions, namely enhancing social connections, providing engaging and meaningful activities, and the affect-generating aspect of the human-animal bond. A fourth mechanism was identified for animal-assisted interventions only: promoting physical activity. Robotic animals appear to have a place in complex human-animal relationships, but a greater understanding of robotic animal interventions is required to harness the benefits that may be derived from their use. Conclusion Delivering these interventions appear promising in improving psychosocial outcomes for people with dementia. As most included studies had methodological limitations, these findings are preliminary, but contribute to the body of evidence providing an understanding in terms of intervention characteristics and mechanisms of action. When developing intervention guidance, attention should be given to potential mechanisms and fundamental characteristics such as session content, delivery format and facilitator role.