This study evaluated the effectiveness of an activity-based intervention (massage) in producing a relaxation response and sleep in nursing home residents. Using a 2 (intervention, control) × 3 (baseline, intervention, withdrawal) mixed design with random assignment to the intervention (n = 15) or control (n = 13) group, results indicated that the massage produced a statistically significant relaxation response as measured by heart rate (p < .001), systolic and diastolic blood pressure (p = .01), and anxiety (p < .001) for individuals receiving the intervention. Analyses had sufficient power despite the sample size. Implications for future research using massage with nursing home residents are discussed.