We assessed the impact and social rating of an active and a passive music condition implemented with six patients with Alzheimer's disease. In the active condition, the patients used a simple hand response and a microswitch to self-regulate music stimulation inputs. In the passive condition, music stimulation was automatically presented throughout the sessions. Active and passive stimulation sessions were preceded and followed by control (non-stimulation) sessions. The active condition sessions showed an increase in the patients' indices of positive participation (e.g., singing or music-related movements, and smiles) greater than that observed in the passive condition sessions for five of the six patients. Positive intervention effects could also spread to the post-intervention sessions. Social raters (42 care and rehabilitation staff members working with persons with multiple disabilities) favored the active condition on a six-item questionnaire dealing with, among others, conditions' suitability, respect of patients' dignity and independence, and practicality. The implications of the findings as to the plausibility/desirability of an active stimulation condition were discussed. (Contains 2 tables and 2 figures.)