The importance of salivary SARS-CoV-2 antibodies, following infection and vaccination, has not been fully established. 875 healthcare workers were sampled during the first wave in 2020 and 66 longitudinally in response to Pfizer BioNTech 162b2 vaccination. We measured SARS-CoV-2 total IgGAM and individual IgG, IgA and IgM antibodies. IgGAM seroprevalence was 39.9%; however, only 34.1% of seropositive individuals also had salivary antibodies. Infection generated serum IgG antibodies in 51.4% and IgA antibodies in 34.1% of individuals. In contrast, the salivary antibody responses were dominated by IgA (30.9% and 12% generating IgA and IgG antibodies, respectively). Post 2nd vaccination dose, in serum, 100% of infection naïve individuals had IgG and 82.8% had IgA responses; in saliva, 65.5% exhibited IgG and 55.2% IgA antibodies. Prior infection enhanced the vaccine antibody response in serum but no such difference was observed in saliva. Strong neutralisation responses were seen for serum 6 months post 2nd-vaccination dose (median 87.1%) compared to low neutralisation responses in saliva (median 1%). Intramuscular vaccination induces significant serum antibodies and to a lesser extent, salivary antibodies; however, salivary antibodies are typically non-neutralising. This study provides further evidence for the need of mucosal vaccines to elicit nasopharyngeal/oral protection. Although saliva is an attractive non-invasive sero-surveillance tool, due to distinct differences between systemic and oral antibody responses, it cannot be used as a proxy for serum antibody measurement.