WHO currently recommends a single dose of typhoid conjugate vaccine (TCV) in high-burden countries based on 2-year vaccine efficacy data from large randomised controlled trials. Given the decay of immunogenicity, the protection beyond 2 years is unknown. We therefore extended the follow-up of the TyVAC trial in Bangladesh to assess waning of vaccine protection to 5 years after vaccination. We conducted a cluster randomised controlled trial (TyVAC; ISRCTN11643110) in Dhaka, Bangladesh, between 2018 and 2021. Children aged 9 months to 15 years were invited to receive a single dose of TCV or Japanese encephalitis vaccine between April 15, 2018, and November 16, 2019, based on the randomisation of their clusters of residence. Children who received the Japanese encephalitis vaccine were invited to receive TCV at the final visit between Jan 6, and Aug 31, 2021, according to the protocol. This follow-on study extended the follow-up of the original trial until Aug 14, 2023. The primary endpoint of this study was to compare the incidence of blood culture-confirmed typhoid between children who received TCV in 2018-19 (the previous-TCV group) and those who received the vaccine in 2021 (the recent-TCV group), to evaluate the relative decline in vaccine protection. We also did a nested study using the test-negative design comparing the recent-TCV and previous-TCV groups with unvaccinated individuals, as well as an immunogenicity study in a subset of 1500 children. Compared with the recent-TCV group, the previous-TCV group had an increased risk of typhoid fever between 2021-23, with an adjusted incidence rate ratio of 3·10 (95% CI 1·53 to 6·29; p<0·0001), indicating a decline in the protection of a single-dose of TCV 3-5 years after vaccination. The extrapolated vaccine effectiveness in years 3-5 was 50% (95% CI -13 to 78), and was validated using the test-negative design analysis, with a vaccine effectiveness of 84% (74 to 90) in the recent-TCV group and 55% (36 to 68) in the previous-TCV group, compared with unvaccinated individuals. Anti-Vi-IgG responses declined over the study period. The highest rate of decay was seen in children vaccinated at younger than 2 years in the original trial. The inverse correlation between age and the decay of antibodies was also seen in the subgroup analysis of vaccine effectiveness, where the youngest age group (<7 years at fever visits) exhibited the fastest waning, with vaccine effectiveness dropping to 24% (95% CI -29 to 55) at 3-5 years after vaccination. A decline in the protection conferred by a single-dose TCV was observed 3-5 years after vaccination, with the greatest decline in protection and immune responses observed in children vaccinated at younger ages. A booster dose of TCV around school entry age might be needed for children vaccinated while younger than 2 years to sustain protection against typhoid fever during the school years when the risk is the highest. The Bill & Melinda Gates Foundation.