This research examined the link between conspiracy beliefs and political decisions in the context of the 2016 European Union referendum in the United Kingdom. In a longitudinal study, we surveyed British participants at two-time points (one week before the referendum vs. immediately after the referendum). At both time points, participants (n = 334) indicated their belief in conspiracy theories specific to the referendum, their general tendencies toward conspiracy theorizing, their political orientation, and support for leaving the EU, followed by how they voted in the referendum, taken at Time 2. Using cross-lagged path analysis, we found that conspiracy beliefs specific to the referendum predicted both support for leaving the EU, and voting to leave the EU, above and beyond political orientation. We also found that the general tendency toward conspiracy theorizing predicted belief in conspiracy theories relevant to the referendum, which subsequently led to increased support for leaving the EU. The chief novelty of this research lies in its longitudinal design, allowing us to conclude that conspiracy beliefs precede political behaviors in a temporal sequence.