The purpose of this study was to evaluate the effect of structured group reminiscence therapy on behavior competence and physical functioning and to compare effectiveness of the intervention by stage of dementia in Taiwanese elders residing in institutional settings. A longitudinal experimental design was used. Seventy-seven Taiwanese elderly participants completed the study; 38 in the intervention group and 39 in the control group. Participants in the intervention group received one reminiscence therapy session weekly for 8 weeks; those in the control group received regular nursing care. Barthel's index (BI) and the Clifton Assessment Procedures for the Elderly Behavior Rating Scale (CAPE-BRS) were administered pre- and post-intervention. Findings indicated that no significant differences between the intervention and control groups on overall behavioral competence (p=0.12) or physical functioning (p=0.44). However, slight overall improvement was identified, statistically significant for the social disturbance subscale of the CAPE-BRS was found (p=0.011). There were no significant changes in overall behavior competence (p=0.11) or physical functioning (p=0.30) by stage of dementia (mild vs. moderate) in the intervention group, counter to the hypothesis. Although our results did not demonstrate a significant effect of structured group reminiscence therapy on overall behavior competency or on physical functioning in elders with dementia, we identified issues related to future implementation of this intervention, such as study power, chosen outcome measures, intervention dose, timing of outcome measures, or cultural applicability issues.