Prior to trial, litigants sometimes conduct broad investigations in which there are multiple opportunities to find supportive evidence by chance alone. During trial, litigants may selectively present only the most helpful evidence uncovered by their investigations. Two experiments examined whether mock jurors appreciate that the evidence they hear at trial may be a selective and unrepresentative sample of underlying facts. The data suggest that people do understand the significance of multiple-opportunity searches for legal inference. However, they may not consider the possibility that evidence was strategically selected from a larger sample space of facts unless that sample space is identified.