The relationship between political judgment and science-based expertise is a troubled one. In the media three cliché images compete. The business-as-usual political story is that, in spite of appearances to the contrary, politics is safely ‘on top’ and experts are still ‘on tap’. The story told by scientists is that power-less but inventive scholars only ‘speak truth to power’. But there is plenty of room for a more cynical interpretation. It sees scientific advisers as following their own interests, unless better (...) paid by other interests, and politicians as asking for advice only to support and legitimize their pre-formed political decisions. To the extent this cynical perspective gains ascendancy, politics and science lose credibility. If we think the three clichés cloak a more complex reality, we should embark upon a quest for other, possibly better models of the science/politics nexus. That is exactly the purpose of this article. Its claim is that a mutual transgression of the knowledge utilization strand of research in policy studies and the study of science, technology and society will provide us with more sophisticated images of science/politics boundary arrangements. Building upon Habermas’ well-known distinctions and Wittrock’s historical-institutional approach in the construction of a property space, eight models are presented. We should try to discover the conditions under which some of these models may claim greater verisimilitude. This may allow us to rethink the role of scientific expertise in policymaking and generate a model that guides experts and policymakers in their day-to-day boundary work. (shrink)