Confirmation bias has been identified as a recurrent cause of wrongful convictions. To mitigate the risk of this bias, police investigators should devote effort to exploring and attempting to falsify alternative hypotheses about a crime. In the Netherlands, investigators in complex capital crime investigations are instructed to apply devil’s advocacy (DA) to increase discussion and testing of alternative investigative theories. This study provides a first quantitative examination of whether DA influences reasoning about evidence in crime investigations. Police investigators who regularly work with DA (n = 63), investigators who do not (n = 76), and laypersons (n = 88) read a fictional investigation report and judged the likelihood that the prime suspect was guilty both before and after asking follow-up questions and receiving additional information. None of the groups clearly demonstrated confirmation bias. Contrary to prediction, experience with DA was not associated with greater falsification efforts. DA-experienced investigators showed a preference for asking falsifying questions only when compared to laypersons and without translating in greater adjustment of guilt-likelihood judgments. Overall, however, falsification efforts were associated with a reduction in guilt-likelihood judgments. Implications for future research and practice are discussed.