Survey-based evidence gathered over the past several decades suggests that substantial minorities of the Canadian public associate immigrants with crime and crime with immigrants. In this note, we ask whether the myth of immigrant criminality imputed to the public is not itself a myth. We question whether the connection is a salient and enduring part of the public’s mindset or whether it is largely an artifact of the closed-ended items employed to explore the topic. We argue that responses to closed-ended questions on this topic are affected by a “halo effect” response bias – a tendency to associated positive attributes with positively evaluated targets and negative attributes to negatively evaluated targets. In support, we show (1) that responses to open-ended questions tell a very different story, (2) that attitudes toward immigrants strongly predict the likelihood of making the immigrant-crime connection when closed-ended items are used, and (3) that priming a possible immigrant-criminal linkage in a survey enhances this likelihood for subsequent items.