Abstract Insufficient descriptions of what comprises animal-assisted interventions and why and how they are presumed to work pose significant challenges to their replication and further scientific development. The purpose of this study was to construct a comprehensive research-based logic model describing a promising program of equine-assisted activities for older adults with Alzheimer’s disease and related dementias. Logic models can advance scientific development of innovative programs by elucidating their critical elements. Employing the qualitative method of directed content analysis, we conducted secondary analyses of previously collected interviews of five providers and fieldnotes of each programmatic session. Employing quantitative direct observational methods, we conducted secondary analysis of eight previously collected hour-long videotapes of four older adults with dementia during the program. The resulting logic model from our secondary analysis of mixed methods data elucidated the program’s purpose, assumptions, programmatic activities, and outcomes, plus congruent linkages across these elements. These findings can guide replication of the program in multiple research and practice contexts and support its future scientific development. Next research steps include evaluating short- and long-term outcomes beyond participants’ direct experiences of the program, clarifying the program’s optimal dosages, and ensuring fidelity of its implementation. This logic model may also help to inform scientific development of other animal-assisted interventions.