Roughly 2 billion people live in areas affected by conflict. Despite this, most behavioral science theories, experiments, and interventions have been created and tested in relatively stable environments. When behavioral scientists do engage in conflict settings, we usually repeat a familiar approach: take what has worked elsewhere, “adapt” it locally, then test it. That approach alone is insufficient. This is an urgent challenge. Conflict is transforming lives around the world on a scale we have never seen before. In 2024, 123 million people were forcibly displaced—the highest figure ever recorded. Civilian deaths in Gaza, Lebanon, and Ukraine have climbed to levels not seen in more than a decade, including among children. I have spent close to ten years applying behavioral science in conflict contexts across 15 countries on 4 continents, and here is the difficult reality: our field still lacks essential insight into how behavioral science operates in the midst of active conflict—and the gaps are larger than many of us assume. At the Neuropaz conference this past February, I was invited to speak about the uncomfortable truths that those working where behavioral science and peacebuilding intersect must face. Below are five of those hard truths, along with what we can do in response. Hard truth #1: We have limited evidence about what actually works in conflict settings Most of the theories, experiments, and tools used in applied behavioral science come from stable environments. Political science has examined the dynamics of conflict; development economics has explored postconflict reconstruction; and social scientists have conducted research in conflict zones…