Federal habeas corpus procedure involves an elaborate set of rules for when state criminal judgments may be reviewed by federal courts. One of these rules— the procedural default rule—forbids federal courts to review state judgments if the state court rejected the proposed claim on procedural grounds. This bar may be overcome by a showing of cause and prejudice (a showing that the outcome of the trial would have been different absent the error). In enforcing this rule, federal courts have failed to realize that there are some claims for which a showing of prejudice is never possible. These claims, sometimes called “structural errors,” are exempt from harmless error review when they arise on direct appeal; in that context, courts have realized that demanding a showing that the error changed the outcome would in many cases be asking the impossible. A particularly troubling example arises when defendants raise Batson claims; a prejudice requirement for such claims would mean that defendants would have to prove that a jury selected in the absence of racial discrimination would have reached a different verdict. Courts have realized that such a showing generally cannot be made, and they have granted relief anyway. What courts have not uniformly acknowledged is that the showing of prejudice is equally impossible in the habeas context. Instead, some courts have decided that prejudice can be presumed for structural errors in this posture, and other courts have required the impossible showing—meaning that such claims are doomed to fail under the procedural default rule in every instance, even where other, less serious claims can sometimes overcome the bar. This Note identifies and traces the two clashing strands of case law—the structural error cases and the procedural default cases—and explicates the underlying incompatibility. It then examines cases that have attempted to adjudicate such claims, and finally proposes a solution: that courts should modify their procedural default test to accommodate these claims.