This Article develops a fresh account of the meaning and constitutional function of the Voting Rights Act’s core provision of nationwide application, Section 2, which has long been portrayed as conceptually opaque, counterproductive in effect, and quite possibly unconstitutional. I argue that Section 2 delegates authority to the courts to develop a common law of racially fair elections, anchored by certain substantive and evidentiary norms, as well as norms about legal change. The central substantive norm is that injuries within the meaning of Section 2 arise only when electoral inequalities owe to race-biased decisionmaking by majority-group actors, whether public or private. As an evidentiary matter, however, plaintiffs need only show a “significant likelihood” of race-biased decisionmaking, rather than proving it more likely than not. So cast (and with a few more details worked out), Section 2 emerges as a constitutionally permissible response to, inter alia, the largely unrecognized problem of election outcomes that are unconstitutional because of the racial basis for the electorate’s verdict—a problem that generally cannot be remedied through constitutional litigation. My account of Section 2 has numerous practical implications. Most importantly, it suggests that electoral arrangements that induce or sustain race-biased voting are vulnerable under Section 2, irrespective of their potentially dilutive effect on minority representation. My account also clears the ground for overruling the many Section 2 precedents that rest on the constitutional avoidance canon, and it helps to resolve a number of prominent circuit splits.