Bivalent BA.1 booster vaccines were offered to adults aged 50 years or older and clinically vulnerable people as part of the 2022 autumn COVID-19 booster vaccination programme in England. Previously, all adults in England had been offered a primary course consisting of two doses of either ChAdOx1-S or monovalent mRNA vaccine and an mRNA monovalent booster vaccine. We aimed to estimate the long-term duration of protection provided by monovalent COVID-19 vaccines, and the incremental vaccine effectiveness of bivalent BA.1 boosters. In this test-negative case-control study, cases of COVID-19 and controls aged 18 years or older were identified from national data for PCR tests done in hospital settings in England. Our analysis was restricted to people with acute respiratory infections coded in the primary diagnosis field. Data for vaccination status were extracted from the English national vaccine register and linked to COVID-19 testing data. Between June 13 and Dec 25, 2022, we estimated the vaccine effectiveness against hospitalisation of two or three or more doses of monovalent COVID-19 vaccines compared with being unvaccinated, stratified by age (18-64 years vs ≥65 years). Between Sept 5, 2022, and Feb 5, 2023, we estimated the incremental vaccine effectiveness (ie, in addition to the protection from earlier vaccines) of receiving a bivalent BA.1 booster vaccine in addition to at least two doses of a monovalent vaccine (when the last dose was at least 6 months ago) among people aged 50 years or older. Analyses were adjusted for week of test, gender, age, COVID-19 risk group, residing in a care home, being a health or social care worker, Index of Multiple Deprivation quintile, ethnicity, and recent COVID-19 positivity. Our analysis of monovalent COVID-19 vaccines included 19 841 cases and 43 410 controls. Absolute vaccine effectiveness against hospitalisation among people who had received at least three doses plateaued from 6 months after the last dose at around 50% in those aged 65 years or older and at around 30% in those aged 18-64 years. Our analyses of the effectiveness of bivalent BA.1 boosters included data for 9954 cases and 39 108 controls aged 50 years or older. Incremental vaccine effectiveness peaked at 53·0% (95% CI 47·9-57·5) 2-4 weeks after administration, before waning to 35·9% (31·4-40·1) after 10 or more weeks. Our study provides evidence that monovalent COVID-19 vaccines offer moderate long-term protection against hospitalisation in people aged 65 years or older and that the bivalent BA.1 booster vaccines were effective in preventing hospitalisation among people aged 50 years or older at a time when omicron lineages were circulating in England. None.