BACKGROUND: Child homelessness is correlated with a wide range of health and psychosocial challenges including poor school performance, juvenile justice involvement, and heightened risk of exposure to early-life violence and trauma. Despite this, participation in therapy tends to be low. Animal Assisted Play TherapyTM (AAPT), a comprehensive model that systematically integrates trained therapy animals into play therapy, serves as a compelling modality for engaging this population into treatment. The tenets of AAPT are aligned with several clinical goal areas that homeless children might address in therapy, including the strengthening of attachment relationships with primary caregivers. METHODS: The study integrated analysis of projective drawings and accompanying narratives with the treatment records of 11 children (ages 6-11) who received canine assisted therapy while residing in an urban, mid-Atlantic family homeless shelter. All children worked with a clinician trained in AAPT and participated in at least three therapy sessions with a qualified therapy dog present. Each child created a drawing in response to the prompt: “Draw a picture of a child and a dog”, and told a story based on the contents of their drawing. Parent/caregivers of each child participated in a qualitative interview that elicited feedback and reflections on the therapy process. Grounded constructivist theory and interpretive description were used to conduct both individual and cross-participant analysis. Analysis was further informed by children’s case history files and parent interviews about children’s developmental history. RESULTS: The projective drawings communicated aspects of homeless children’s relationships with the therapy dog and, in turn, with their primary caregivers. Developmentally, children drew at lower levels than would be expected for their age. Each child personalized their drawing, either by identifying the protagonist as their gender, or including a physical characteristic (e.g. clothing, hairstyle) unique to the child; this suggests that the children tapped into their personal experience. Several themes emerged from analysis of the drawings and narratives including representations of lived and wished-for attachment experiences. Children depicted relationships between the characters in their drawings and narratives that were characterized by emotional closeness as well as frequent separations and reunions. Children also highlighted the importance of learning tasks related to training and caring for the dogs. These themes were reflected in the children's treatment records and the parent/caregiver interviews. Parent/caregivers described their child's experience in therapy positively, and identified the therapy dog as a component of the treatment's success. IMPLICATIONS: Projective drawings enabled homeless children to communicate their attachment experiences in a manner sensitive to their developmental needs. Themes that emerged from this study inform further research on specific benefits of animal assisted therapy. Specifically, the themes of lived and wished-for attachment experiences suggest that further research on this modality might focus on the ways that the modality allows children to build new relationships and strengthen existing ones. The drawings created during this study are a valuable tool in understanding the experiences of homeless children, and lay the groundwork for further study of the use of projective drawings for exploring children's experiences in therapy.