Community engaged (CE) methods are used in health-related research, but few discuss methodological aspects among older adults. We describe the methodology and lessons learned from a CE study of whether ARTmail, a structured participatory arts program, benefited older adults aged 60+ with memory symptoms / cognitive impairment (MS/CI). Our study, conducted in 2015-2017 with support from the NEA, was a partnership between a non-profit organization in North Carolina that provides creative programming for older adults with varied abilities, and researchers in an area University. Older adults with MS/CI symptoms receiving care in area communities were recruited into either the art intervention or a control group, in collaboration with community staff. We reflect on the CE research process with a partnership among community organizations, researchers, care staff, and older adults. Asymmetries in priorities, resources, and decision-making power are described, and implications for the research process and findings are discussed.