Motivated beliefs refer to situations where people process information in a way which is favorable for themselves: they believe what they want to believe (to some extent). Although laboratory experiments have tested this phenomenon extensively, there is a stark lack of evidence from the field. In this study, we exploit the opportunity of a natural experiment which happened in the U.K. during the COVID-19 vaccination campaign. UK residents could choose to opt into the vaccination program, but not which vaccine they received. Once age and health restrictions are taken into account, the assignment to a specific vaccine was, thus, random. In the Summer 2021, we interviewed over 1,000 non-vaccinated U.K. residents about their beliefs on the characteristics of different COVID-19 vaccines. A few months later, we reinterviewed the same respondents, when most of them had been vaccinated. Using a difference-in-differences estimator, we show that people positively updated their beliefs about the vaccine they received, far more than people who got another vaccine (or no vaccine at all). The effect seems to be driven by those who received a different vaccine than the one they preferred. These findings corroborate the existence of motivated beliefs in the field, even when beliefs relate to unchosen outcomes.