Gordon Kaufman's “constructive theology” can easily be taken out of context and misunderstood or misrepresented as a denial of God. It is too easily overlooked that in his approach everything is an imaginary construct given no immediate ontological status—the self, the world, and God are “products of the imagination.” This reflects an influence, not only of theories on linguistic and cultural relativism, but also of Kant's “ideas of pure reason.” Kaufman is explicit about this debt to Kant. But I argue there are other aspects of Kant's legacy implicit in his method. These center around Kaufman's engagement with “observed patterns” in nature. With Paul Tillich's aid, I bring this neglected issue to the fore and argue that addressing it allows one to more readily capitalize upon the Kantian influence in Kaufman's method. This, in turn, encourages one to tap more deeply into the epistemic underpinnings of Kaufman's approach to the science–religion dialogue.