Aims: To elucidate varicella zoster virus (VZV)-specific cell-mediated immunity and humoral immunogenicity against live attenuated Oka varicella zoster vaccine concurrently vaccinated with 23-valent pneumococcal polysaccharide vaccine (PPSV23) in elderly people with diabetes mellitus. Methods: This double-blind randomized controlled single-centre study of 60–70-year-old people with diabetes compared immunity and safety profiles 3 months after one dose of varicella zoster vaccine or placebo. PPSV23 was immunized simultaneously. Primary analysis evaluated cell-mediated immunity using the VZV skin test. Secondary analyses were a VZV interferon–γ enzyme-linked immunospot (ELISPOT) assay and immunoadherence haemagglutination test. Adverse experiences were recorded using diary questionnaires. Results: By intent-to-treat analysis, 27 participants with diabetes who had been administered the vaccine were compared with 27 participants who were given a placebo. Changes in skin test scores were 0.41 ± 0.80 and 0.11 ± 0.93 (P = 0.2155), and geometric mean fold rises of the ELISPOT counts were 1.2 [95% confidence interval (CI) 0.2, 7.9] and 1.2 (95% CI 0.2, 7.3) (P = 0.989) in the vaccine and placebo groups, respectively. The geometric mean titre did not increase 3 months after vaccination in either group. No vaccination-related severe adverse experience was reported and no participant developed herpes zoster. Discussion: Our previous results demonstrated that varicella zoster vaccine safely enhanced VZV-specific immunity in elderly people with or without diabetes. The results of this study showed that varicella zoster vaccine can be used safely, but it cannot boost virus-specific immunity in elderly people with diabetes when administered with concurrent PPSV23. Alternative strategies are needed to prevent VZV-associated diseases in this population.