A cost-utility analysis (CUA) was applied to group living for dementia patients. A Markov-model of an expected life-length of 8 years was used. Forty-six patients in group living were compared to 39 patients living at home by inclusion and 23 institutionalized patients. When the cost per gained quality-adjusted life-year (QALY) was calculated, the group living alternative was the most favorable for the patients, giving a cost per paired QALY of US dollars > 0. In the extensive sensitivity analysis the main result was consistent but methodological problems were indicated.