SummaryBackgroundAntibody responses to SARS-CoV-2 vaccination vary for reasons that remain poorly understood.MethodsWe tested for presence of combined IgG, IgA and IgM (IgGAM) anti-spike antibodies before and after administration of two doses of ChAdOx1 nCoV-19 (ChAdOx1, Oxford-AstraZeneca) or BNT162b2 (Pfizer-BioNTech) in UK adults participating in a population-based longitudinal study who received their first dose of vaccine from December 15, 2020 to July 10, 2021. Information on sixty-six potential sociodemographic, behavioural, clinical, pharmacological and nutritional determinants of serological response to vaccination was captured using serial online questionnaires. We used logistic regression to estimate multivariable-adjusted odds ratios (aORs) for associations between independent variables and risk of seronegativity following two vaccine doses. Participants who were seronegative after receiving two vaccine doses were offered an additional antibody test following subsequent administration of a ‘booster’ dose of BNT162b2 or mRNA-1273 (Moderna) from September 23 to December 12, 2021.FindingsSerology results following two vaccine doses were available for 9,101 participants, of whom 5,770 (63.4%) received ChAdOx1 and 3,331 (36.6%) received BNT162b2. Anti-spike IgGAM was undetectable in 378 (4.2%) participants at a median of 8.6 weeks (IQR 6.4-10.7 weeks) after their second dose of vaccine. Seronegativity following two doses of SARS-CoV-2 vaccination was associated with administration of ChAdOx1 vs BNT162b2 (aOR 7.03, 95% CI 4.39-11.24), shorter interval between first and second vaccine doses (aOR 2.37, 1.06-5.26, for <6 weeks vs >10 weeks; aOR 1.59, 1.18-2.13, for 6-10 weeks vs >10 weeks), poorer self-assessed general health (aOR 3.33, 1.49-7.46, for poor vs excellent), immunodeficiencies (aOR 6.75, 2.63-17.35) and prescription of systemic immunosuppressants (aOR 3.76, 2.44-5.78). By contrast, pre-vaccination SARS-CoV-2 seropositivity (aOR 0.16, 0.04-0.70, for symptomatic seropositives vs seronegatives) and supplemental vitamin D intake (aOR 0.73, 0.53-0.99) were associated with reduced risk of post-vaccination seronegativity. 247/378 (65.3%) of participants who were seronegative after two doses of ChAdOx1 vs BNT162b2 provided a third sample at a median of 7.8 weeks (IQR 5.8-10.4) after receiving a booster dose of BNT162b2 or mRNA-1273: eight (3.2%) of them remained seronegative after three vaccine doses, all of whom either had a primary immunodeficiency or were taking systemic immunosuppressant drugs.InterpretationWe identify multiple determinants of antibody responses to two doses of ChAdOx1 or BNT162b2, many of which are potentially modifiable. Booster doses of BNT162b2 or mRNA-1273 were highly effective in achieving seroconversion in those who failed to mount antibody responses following two doses of ChAdOx1 or BNT162b2.Study registrationhttps://clinicaltrials.gov/ct2/show/NCT04330599FundingBarts Charity, Fischer Family Trust, The Exilarch’s Foundation, DSM Nutritional Products, Health Data Research UKResearch in contextEvidence before this studyWe searched PubMed, medRxiv, and Google Scholar for papers published from January 1, 2020, to February 1, 2022, using the search terms (antibody OR humoral OR serologic* OR immunogenic*) AND (SARS-CoV-2 vaccine OR ChAdOx1 or BNT162b2 coronavirus), with no language restrictions. Population-based studies investigating multiple potential determinants of vaccine immunogenicity in people with known pre-vaccination SARS-CoV-2 serostatus are lacking.Added value of this studyThis large population-based study, conducted in a population with known pre-vaccination SARS-CoV-2 serostatus, examines a comprehensive range of potential sociodemographic, behavioural, clinical, pharmacological and nutritional determinants of antibody responses to administration of two major SARS-CoV-2 vaccines (i.e., ChAdOx1 or BNT162b2), many of which have not previously been investigated. It is also the first population-based study to characterise antibody responses to booster doses of SARS-CoV-2 vaccines in adults who were seronegative after their primary course of vaccination.Implications of all the available evidenceIncreased risk of seronegativity following two doses of SARS-CoV-2 vaccines was associated with administration of ChAdOx1 vs BNT162b2, shorter interval between first and second vaccine doses, poorer self-assessed general health, immunocompromise and SARS-CoV-2 seronegativity pre-vaccination. Regular intake of vitamin D supplements was associated with reduced risk of post-vaccination seronegativity. Randomised controlled trials are now needed to test for causality. Booster doses of BNT162b2 or mRNA-1273 were highly effective in achieving seroconversion in the majority of people who failed to mount antibody responses following a primary course of vaccination, the few exceptions being a subset of those with primary immunodeficiency or systemic immunosuppressant drugs.