We introduce the first set of stimuli designed to resolve methodological and theoretical issues that have muddled the interpretation of results on the memorability of supernatural concepts (e.g., ghosts, souls, spirits), an important line of research in the cognitive science of religion (Barrett, 2007). We focus here on Boyer’s (1994), Boyer, 2001) pioneering minimally counterintuitive (MCI) hypothesis according to which supernatural concepts tap a special memory-enhancing mechanism linked to violations of default intuitive inferences. Empirical tests of the MCI account have given rise to a vexed picture that renders meaningful interpretation difficult. The lack of a common standard of comparison among different studies, coupled with the presence of uncontrolled variables independently known to affect memorability, lie at the heart of these problems. We show that our new stimuli offer the hope of resolving these issues, thereby establishing a more secure foundation for the study of the memorability of supernatural concepts.