BackgroundThe extended use of varicella vaccine in adults aged 50 and older against herpes zoster (HZ) was recently approved in Japan, which has raised the need to evaluate its value for money. MethodsWe conducted a cost-effectiveness analysis with Markov modelling to evaluate the efficiency of varicella vaccine immunisation programme for the elderly in Japan. Four strategies with different ages to receive a shot of vaccine were set, namely: (1) 65–84, (2) 70–84, (3) 75–84 and (4) 80–84years old (y.o.). Incremental cost-effectiveness ratios (ICERs) compared with no programme from societal perspective were calculated. The health statuses following the target cohort are as follows: without any HZ-related disease, acute HZ followed by recovery, post-herpetic neuralgia (PHN) followed by recovery, post HZ/PHN, and general death. The transition probabilities, utility weights to estimate quality-adjusted life year (QALY) and disease treatment costs were either calculated or cited from literature. Costs of per course of vaccination were assumed at ¥10,000 (US$91). The model with one-year cycle runs until the surviving individual reached 100 y.o. ResultsICERs ranged from ¥2,812,000/US$25,680 to ¥3,644,000/US$33,279 per QALY gained, with 65–84 y.o. strategy having the lowest ICER and 80–84 y.o. strategy the highest. None of the alternatives was strongly dominated by the other, while 80–84 y.o. and 70–84 y.o. strategy were extendedly dominated by 65–84 y.o. strategy. Probabilistic sensitivity analyses showed that the probabilities that ICER is under ¥5,000,000/US$45,662 per QALY gained was at 100% for 65–84 y.o., 70–84 y.o., 75–84 y.o. strategy, respectively, and at 98.4% for 80–84 y.o. strategy. ConclusionWe found that vaccinating individuals aged 65–84, 70–84, 75–84, and 80–84 with varicella vaccine to prevent HZ-associated disease in Japan can be cost-effective from societal perspective, with 65–84 y.o. strategy as the optimal alternative. Results are supported by one-way sensitivity analyses and probabilistic sensitivity analyses.