Background: Discontinuation of benzodiazepine usage has never been evaluated in economic terms. This study aimed to compare the relative costs and outcomes of tapering off long-term benzodiazepine use combined with group cognitive behavioural therapy (TO+CBT), tapering off alone (TOA) and usual care. Method: A randomised controlled trial was conducted, incorporating a costeffectiveness analysis from a societal as well as a pharmaceutical perspective. The cost of intervention treatment, prescribed drugs, healthcare services, productivity loss, and patients' costs were measured using drug prescription data and cost diaries. Costs were indexed at 2001 prices. The principal outcome was the proportion of patients able to discontinue benzodiazepine use during the 18-month follow-up. A secondary outcome measure was quality of life (Health Utility Index Mark III [HUI-3] and the Medical Outcomes Study 36-item Short-Form Health Survey [SF-36]). Results: A total of 180 patients were randomised to one of TO+CBT (n = 73), TOA (n = 73) or usual care (n = 34). Intervention treatment costs were an average of €172.99 per patient for TO+CBT and €69.50 per patient for TOA. Both treatment conditions significantly reduced benzodiazepine costs during follow-up compared with usual care. The incremental cost-effectiveness ratios (ICERs) showed that, for each incremental 1% successful benzodiazepine discontinuation, TO+CBT cost €10.30-62.53 versus usual care, depending on the study perspective. However, TO+CBT was extendedly dominated or was dominated by TOA. This resulted in ICERs of €0.57, €10.21 and €48.92 for TOA versus usual care from the limited pharmaceutical, comprehensive pharmaceutical and societal perspective, respectively. Conclusions: TO+CBT and TOA both led to a reduction in benzodiazepine costs. However, it remains uncertain which healthcare utilisation has a causal relationship with long-term benzodiazepine consumption or its treatment. Although the ICERs indicated better cost effectiveness for TOA than for TO+CBT, the differences were relatively small. The addition of group CBT to tapering off had no clinical or economic advantages. Extrapolation of our data showed that the investment in TOA was paid back after 19 months when corrected for treatment gain with usual care.