Parkinson's disease (PD) is a progressive neurodegenerative disease associated with deficits in motor, cognitive, and emotion/quality of life (QOL) domains, yet most pharmacologic and behavioral interventions focus only on motor function. Our goal was to perform a pilot study of Dance for Parkinson's-a community-based program that is growing in popularity-in order to compare effect sizes across multiple outcomes and to inform selection of primary and secondary outcomes for a larger trial. Study participants were people with PD who self-enrolled in either Dance for Parkinson's classes (intervention group, N=8) or PD support groups (control group, N=7). Assessments of motor function (Timed-Up-and-Go, Gait Speed, Standing Balance Test), cognitive function (Test of Everyday Attention, Verbal Fluency, Alternate Uses, Digit Span Forward and Backward), and emotion/QOL (Geriatric Depression Scale, Falls Efficacy Scale-International, Parkinson's Disease Questionnaire-39 (total score and Activities of Daily Living subscale)) were performed in both groups at baseline and follow-up. Standardized effect sizes were calculated within each group and between groups for all 12 measures. Effect sizes were positive (suggesting improvement) for all 12 measures within the intervention group and 7 of 12 measures within the control group. The largest between-group differences were observed for the Test of Everyday Attention (a measure of cognitive switching), gait speed and falls efficacy. Our findings suggest that dance has potential to improve multiple outcomes in people with PD. Future trials should consider co-primary outcomes given potential benefits in motor, cognitive and emotion/QOL domains.