Summary Background The licensed live, attenuated varicella-zoster virus vaccine prevents herpes zoster in adults older than 50 years. We aimed to determine whether intradermal administration of zoster vaccine could enhance vaccine immunogenicity compared with conventional needle subcutaneous administration. Methods In this randomised, dose-ranging study, adults aged 50 years or older who had a history of varicella or who had resided in a country with endemic varicella-zoster virus infection for 30 years or more were eligible. Participants received the approved full or a 1/3 dose of zoster vaccine given subcutaneously or one of four intradermal doses (full, 1/3, 1/10, or 1/27 dose) using the MicronJet600 device. The two subcutaneous doses and the four intradermal doses were randomised (1·5:1:1:1:1:1) by computer generated sequence with randomisation stratified by age (50–59 years or 60 years or older). The primary immunogenicity endpoint was the change from baseline in IgG antibody to varicella-zoster virus-specific glycoproteins (gpELISA) measured at 6 weeks. All patients were included in the primary and safety analyses. This study is registered with ClinicalTrials.gov , number NCT01385566. Findings Between Sept 2, 2011, and Jan 13, 2012, 224 participants were enrolled from three clinics in the USA and 223 were randomly assigned: 52 to receive the full dose subcutaneous zoster vaccine, 34 to receive the 1/3 dose subcutaneous zoster vaccine, 34 to receive the full dose intradermal zoster vaccine, 35 to receive the 1/3 dose intradermal zoster vaccine, 34 to receive the 1/10 dose intradermal zoster vaccine, and 34 to receive the 1/27 dose intradermal zoster vaccine. Full dose zoster vaccine given subcutaneously resulted in a gpELISA geometric mean fold-rise (GMFR) of 1·74 (90% CI 1·48–2·04) at 6 weeks post-vaccination compared with intradermal administration which resulted in a significantly higher gpELISA GMFR of 3·25 (2·68–3·94; p<0·0001), which also remained high at 18 months. An apparent dose–response relation was observed with intradermal administration (1/3 dose subcutaneous GMFR 1·64 90% CI 1·36–1·99, 1/3 dose intradermal 2·58 (2·13–3·13), 1/10 dose intradermal 2·22 1·83–2·69, and 1/27 dose intradermal 1·64 1·35–2·00). Each partial dose of zoster vaccine given intradermaly had a gpELISA GMFR comparable to that of full dose zoster vaccine given subcutaneously. Transient erythema and induration were more common after intradermal administration (31% erythema for full subcutaneous dose and 77% for intradermal dose). Interpretation Intradermal zoster vaccine showed a greater increase in varicella-zoster virus gpELISA antibody compared with subcutaneous zoster vaccine at comparable doses. Larger and longer studies of intradermal administration of live, attenuated zoster vaccine are needed to provide convincing evidence of improved cell mediated immunity. Funding Merck & Co Inc.