Forensic evidence plays a critical role in court proceedings and the administration of justice. It is a powerful tool that can help convict the guilty and avoid wrongful conviction of the innocent. Unfortunately, flaws in forensic evidence are increasingly becoming apparent. Assessments of forensic science have too often focused only on the data and the underlying science, as if they exist in isolation, without sufficiently addressing the process by which forensic experts evaluate and interpret the evidence. After all, it is the forensic expert who observes the data and makes interpretations, and therefore forensic evidence is mediated by human and cognitive factors. A U.S. National Research Council examination of forensic science in 2009, followed by a 2016 evaluation by a presidential panel, along with a U.K. inquiry into fingerprinting in 2011 and a 2015 guidance by the U.K. Forensic Science Regulator, have all expressed concerns about biases in forensic expert decision-making. Where does forensic bias come from, and how can we minimize it?