The transcendental logic in Kant’s Critique of Pure Reason is famously divided into a “logic of truth” and a “logic of illusion”, a logic of correct thinking and a logic of false thinking: a transcendental analytic and a transcendental dialectic. That division is peculiar to transcendental logic, as pure general (or formal) logic is simply a logic of (formal) truth. Why does transcendental logic need to include a separate logic of illusion? Why do the rules of (transcendental) truth not suffice in order to determine the opposite of truth? For Kant, reason falls into transcendental illusion not by violating, but by following its very own rules. How should we make sense of that conception? The paper explores the relation between the “logical maxim” and the transcendental “principle” of reason, as well as the way Kant resolves the antinomy of reason, in order to argue that the critique of transcendental illusion does indeed require the unfolding of the “logic of illusion” and cannot be a corollary to the “logic of truth”. Besides the “canon” of the proper use of the understanding and the power of judgement, transcendental logic must include a separate “discipline” of reason.