OBJECTIVES: To assess the cost-effectiveness of memantine in moderate and severe AD patients who exhibit agitation/aggression and psychotic symptoms (APS) from the UK National Health Service and Personal Social Services perspective. METHODS: The cost-utility analysis was based on 5-year Markov cohort simulations. The model evaluated the impact of memantine on time to Full-Time-Care (FTC), Quality-Adjusted-Life-Years (QALYs) and costs, in pre-FTC patients compared with standard care, i.e. no pharmacotherapy or background treatment with acetylcholinesterase inhibitors. FTC was defined based on locus of care and patient’s physical and functional dependency status. Transition probabilities, baseline characteristics, resource utilization volumes, health utility weights and mortality rates were derived from the 4.5-year London and South-East Region (LASER-AD) epidemiological study. Effectiveness estimates came from a meta-analysis of six large randomised clinical trials. Costs covered routine patient management, hospitalization, social community services, institutionalisation, and medications. Results were reported in EUR (GBP), 2009. The model underwent extensive stochastic and one-way sensitivity analyses, testing the model assumptions and changes in input parameters. RESULTS: Over five years, patients receiving standard care spend on average 78.8 weeks in the pre-FTC state. Overall costs in this group were €117,960 (£98,810). QALYs were estimated at 1.49 (30% of full health). Memantine was associated with a longer time-to-FTC of 11.2 weeks, QALY gains of 0.07 and cost-savings of €5930 (£4970). Lower costs in the memantine group were due to prolonged pre-FTC period. Memantine was more effective and less costly strategy relative to standard care in 99.98% of simulations. The estimated benefits and cost savings were almost twice higher than those previously estimated in all moderate and severe AD patients, largely due to enhanced efficacy of memantine in APS patients, who, when left untreated, rapidly deteriorate. CONCLUSIONS: The model showed that memantine yielded higher benefits at no additional costs relative to its alternative.