We assess whether the COVID-19 vaccine induces COVID-19 risky behavior (e.g., going to bars and restaurants) and thus reduces vaccine efficacy. A key empirical challenge is the endogeneity bias when comparing risk-taking by vaccination status since people choose whether to get vaccinated. To address this bias, we exploit rich survey panel data on individuals followed before and after vaccine availability over 14 months in an event study fixed effects model with individual, time, sector, and county-by-time fixed effects and inverse propensity weights. We find evidence that vaccinated persons, regardless of the timing of vaccination, increase their risk-taking by increasing engagement in some risk-taking activities. The evidence is consistent with the "lulling effect". While vaccine availability may reduce the risk of contracting COVID-19, it also contributes to further spread of the virus by incentivizing risk-taking in the short term.