Abstract: This paper studies the direct and indirect effectiveness of Covid-19 vaccines among vaccinated healthcare workers and their unvaccinated adult household members in a mass vaccine program in Finland. Methods: We used national databases that record all polymerase chain reaction (PCR)-confirmed SARS-CoV-2 infections and mRNA-based (BNT162b2 by Pfizer-BioNTech or mRNA-1273 by Moderna) vaccine doses administered in Finland since the beginning of the epidemic. These data were merged with administrative full population datasets that include information on each person’s occupation and unique identifiers for spouses living in the same household. To estimate the direct and indirect effectiveness of mRNA-based vaccines in a household setting, we compared the cumulative incidence of PCR-confirmed SARS-CoV-2 infections between vaccinated and unvaccinated healthcare workers as well as between their unvaccinated spouses. Findings: Our estimates imply indirect effectiveness of 8.7% (95% CI: -28.9 to 35.4) two weeks and 42.9% (95% CI: 22.3 to 58.1) 10 weeks after the first dose. The effectiveness estimates for unvaccinated household members are substantial, but smaller than the direct effect and occur more gradually among unvaccinated household members than among vaccinated individuals. Interpretation: Our results suggest that mRNA-based vaccines do not only prevent SARS-CoV-2 infections among vaccinated individuals but lead to a substantial reduction in infections among unvaccinated household members. The results are consistent with the notion that mRNA-based vaccines affect susceptibility in vaccinated individuals and prevent transmission from vaccinated to unvaccinated individuals.