Those who wish to defend the role of aesthetic values in science face a dilemma: Aesthetic language is either used metaphorically for what are ultimately epistemic features, or it is used literally, but the importance of such values for science are unclear. This paper introduces a new account that gets around this problem by looking to an overlooked source of aesthetic value in science: the relation between form and content. A fit between the content of a thought experiment, and the way in which that content is formulated, can have important epistemic pay offs through contributing to scientific understanding.