We develop, validate, and demonstrate the value of an experimental paradigm for studying exposure to fake news. This paradigm complements existing methods for studying fake news by helping researchers address questions that current methods cannot. It allows participants to be ethically exposed to real and fake news items. These news items, like their real-world counterparts, report information that is consequential for later (in-experiment) decisions. But unlike their real-world counterparts, these news items are generated within a news ecosystem existing solely within the context of the experiment (and thus under the control of the experimenter). Through this paradigm, we provide a way to manipulate participants’ exposure to fake news without misinforming them (relative to a control group). This makes it possible to study the effect of mere exposure to fake news on participants, rather than a combined effect of exposure and misinformation. In two sets of preregistered experiments, we establish the validity of the manipulation, and show how exposure to fake news changes how much people are willing to pay for real news.