Conventional serological approaches lack sensitivity for the detection of recent SARS-CoV-2 infections in vaccinated individuals, as these individuals exhibit a blunted anti-nucleocapsid (N) response. This limitation was recently addressed by the development of a "ratio-based approach", which compares longitudinally collected specimens. Here, we used this approach to estimate the incidence of SARS-CoV-2 infection and reinfection in Québec (Canada) during the Omicron wave. Consenting plasma donors were included if they donated plasma before December 15, 2021 and during six consecutive periods of ~ 3months between December 15, 2021 and July 7, 2023 (study period). Anti-N levels were measured with an enzyme-linked immunosorbent assay, and seroconversion was characterized by a ratio of ≥ 1.5 between the optical density of two consecutive samples. Among the 254 donors, the adjusted proportion of donors (95% confidence interval [CI]) with a new infection ranged between 18.1% (13.2‒23.0) and 24.2% (18.8‒29.7) over Periods 1-5 and fell to 7.9% (4.9‒11.0) during Period 6. During the study period, the proportion of newly infected donors decreased among those aged < 60 (Period 1 = 31.6%, Period 5 = 4.4%), but increased among those aged ≥ 70 (Period 1 = 0.3%, Period 6 = 10.3%). Throughout the study period, 72 (28.3%) reinfections occurred, including two seroconversion events in a single donor. Overall, 87.4% (95% CI = 82.7‒91.2) were infected by SARS-CoV-2 at least once during the study period. The vast majority of the Québec population may have been infected during the Omicron wave. This longitudinal survey demonstrates the usefulness of the "ratio-based approach" for identifying both new infections and reinfections in a vaccinated population.