From the Agency Detection Device (ADD) to Theory of Mind (ToM), the cognitive faculties involved in the production and sustenance of religious belief are well known: regardless of the extra-mental reality or otherwise of the objects of religious belief, to have such belief is natural and normal, across a wide range of definitions of these terms. Unbelief, however, has not received nearly so much attention, a fact that may be due to the explicit or implicit treatment of religious belief as a deviation from a state of natural cognitive innocence, comparable to a malfunction or mind-disease, or a deficiency incompatible with rational human perfection, once superstition has been banished. Paying careful attention to range of positions that could be characterized as 'atheist', I summarize new research that has examined cognitive capacities or deficiencies that correlate with a professed stance of unbelief. Is atheism, then, rational? Are atheists 'normal'?