OBJECTIVES: Mass COVID-19 immunization campaigns altered the pandemic's progress by protecting the vaccine recipient and reducing transmission. However, evidence for indirect vaccine effectiveness (IVE) is limited due to the difficulties of ascertaining this type of protection. METHODS: Using linked national Brazilian databases, we adapted the test-negative design to evaluate the IVE against symptomatic infection. We analyzed data from January 1 to December 1, 2021 (pre-omicron) and January 1 to April 30, 2022 (omicron BA.1 and BA.2). We compared the probability of testing positive across various levels of second ancestral-strain monovalent COVID-19 vaccine dose coverage, including only unvaccinated individuals in the main analysis and both vaccinated and unvaccinated individuals in additional analyses. Sensitivity analysis focused on children younger than 12 years who did not have access to COVID-19 vaccines during the pre-omicron period. RESULTS: We included 11,039,315 unvaccinated individuals tested during the pre-omicron study period. IVE was minimal until 30% vaccination coverage (<10%), then it followed a dose-dependent pattern, peaking at 37.7 (95% confidence interval 32-42.8) at 70% coverage. For children younger than 12 years, IVE peaked at 59.8% (95% confidence interval 52.7-65.9) at 70% coverage. During the omicron period, IVE remained constant at about 5% across all comparisons. CONCLUSIONS: Our findings confirm that high vaccination coverage using vaccines that prevent infection indirectly protects the community. However, IVE was substantially higher during the pre-omicron period.