Abstract Background Dementia is a progressive disease that decreases quality of life of persons with dementia and is associated with high societal costs. The burden of caring for persons with dementia also decreases the quality of life of family caregivers. The objective of this study was to assess the societal cost-effectiveness of Namaste Care Family program in comparison with usual care in nursing home residents with advanced dementia. Methods Nursing homes were randomized to either Namaste Care Family program or usual care. Outcome measures of the cluster-randomized trial in 231 residents included Quality of Life in Late-Stage Dementia (QUALID) and the Gain in Alzheimer Care Instrument (GAIN) for family caregivers over 12 months of follow-up. Health states were measured using the EQ-5D-3L questionnaire which were translated into utilities. QALYs were calculated by multiplying the amount of time a participant spent in a specific health state with the utility score associated with that health state. Healthcare utilization costs were estimated using standard unit costs, while intervention costs were estimated using a bottom-up approach. Missing cost and effect data were imputed using multiple imputation. Bootstrapped multilevel models were used after multiple imputation. Cost-effectiveness acceptability curves were estimated. Results The Namaste Care Family program was more effective than usual care in terms of QUALID (− 0.062, 95%CI: − 0.40 to 0.28), QALY (0.0017, 95%CI: − 0.059 to 0.063) and GAIN (0.075, 95%CI: − 0.20 to 0.35). Total societal costs were lower for the Namaste Care Family program as compared to usual care (− 552 €, 95%CI: − 2920 to 1903). However, these differences were not statistically significant. The probability of cost-effectiveness at a ceiling ratio of 0 €/unit of effect extra was 0.70 for the QUALID, QALY and GAIN. Conclusions The Namaste Care Family program is dominant over usual care and, thus, cost-effective, although statistical uncertainty was considerable. Trial registration Netherlands Trial Register (http://www.trialregister.nl/trialreg/index.asp, identifier: NL5570, date of registration: 2016/03/23).