High-profile political endorsements by scientific organizations and publications have been common in recent years, which raises concerns about political backlash against the endorsing organization in particular and scientific expertise in general. In a pre-registered large-sample online controlled experiment, I randomly assign subjects to receive information about Nature’s endorsement of Joe Biden for the 2020 U.S. presidential election in the context of the COVID-19 pandemic. Seeing the endorsement message leads to large reductions in self-reported trust in Nature among Trump supporters. This distrust translates into lower demand for COVID-related information provided by Nature, as evidenced by substantially reduced requests for Nature articles on vaccine efficacy and emerging COVID variants when offered. The endorsement also reduces Trump supporters’ stated trust in scientists in general, suggesting “reputational externalities” on the scientific community. Treatment effect estimates for Biden supporters are positive but small and often statistically insignificant across outcomes. Finally, I find little evidence that the endorsement is successful at changing subjects’ views about Biden and Trump. These results show political endorsement by scientific organizations can undermine public confidence in the endorsing organization and in the scientific community.