<p>Scientific institutions increasingly engage in politics through candidate endorsements and public advocacy, raising questions about the neutrality of scientific institutions and public trust in science. Analysis of Nature’s presidential endorsements and experimental studies demonstrates that political involvement decreases perceived scientific credibility. While politicized science can still meet the demands of scientific objectivity, the trust costs are substantial. This study examines arguments that political involvement by scientists is unavoidable, finding these justifications insufficient except in rare cases where greater scientific benefits are accrued through political involvement than the corresponding loss of public trust. It concludes that scientists should generally avoid political involvement to preserve institutional credibility and public support for scientific research. Policy implications include establishing ethics review processes for institutional political engagement and improving science education to help the public distinguish between scientific expertise and political advocacy.</p>