BACKGROUND: People with dementia report lower quality of life, but we know little about what interventions might improve it. METHODS: We systematically reviewed 20 randomized controlled trials reporting the effectiveness of non-pharmacological interventions in improving quality of life or well-being of people with dementia meeting predetermined criteria. We rated study validity with a checklist. We contacted authors for additional data. We calculated standardized mean differences (SMD) and, for studies reporting similar interventions, pooled standardized effect sizes (SES). RESULTS: Pooled analyses found that family carer coping strategy-based interventions (four studies, which did not individually achieve significance; n = 420; SES 0.24 (range 0.03-0.45)) and combined patient activity and family carer coping interventions (two studies, not individually significant; n = 191; SES 0.84 (range 0.54-1.14)) might improve quality of life. In one high-quality study, a care management system improved quality of life of people with dementia living at home. Group Cognitive Stimulation Therapy (GCST) improved quality of life of people with dementia in care homes. CONCLUSION: Preliminary evidence indicated that coping strategy-based family carer therapy with or without a patient activity intervention improved quality of life of people with dementia living at home. GCST was the only effective intervention in a higher quality trial for those in care homes, but we did not find such evidence in the community. Few studies explored whether effects continued after the intervention stopped. Future research should explore the longer-term impact of interventions on, and devise strategies to increase, life quality of people with dementia living in care homes or at home without a family carer.