Good faith is a protean concept: it takes on different shades of meaning, and has been made to do many quite distinctive things, in private, public, and international law. These qualities have made good faith perplexing. Partly in consequence, conventional wisdom on good faith is decidedly anti-theoretical. Some have argued that good faith is indefinable. Others argue that the law knows no concept of good faith: reference to good faith is just a placeholder for its antonym, bad faith. In this essay, I suggest that the conventional wisdom is wrong. One can extract a coherent, generalizable, notion of good faith from the law. On the theory presented here, illustrated in respect of voluntary obligations, good faith consists in personal integrity with respect to juridical speech acts. To evince good faith in this sense is to demonstrate fidelity to the known legal significance (objective legal meaning) of one's speech acts and attendant jural relations. A person who acts in good faith behaves righteously in a specifically legal sense. Bad faith, correlatively, involves lack of integrity: iniquity. Iniquity implies a defect in one's deliberate behavior, and hence goes to character, consistent with the intuition that bad faith is morally culpable. But bad faith in law has distinctively legal upshots: it threatens to undermine the institutional integrity of law. Thus, good faith is properly considered an über-norm of foundational importance to law and to legality.