There have been persistent philosophical efforts to demarcate the province of science. Fewer attempts have been made to explore whether these demarcation strategies are consistent with the liberal promise of religious neutrality (LPRN). Within this framework, most liberal political theorists seem to agree that hypotheses suggesting supernatural agency should remain outside the purview of science by principle. In their view, this rule of methodological naturalism (MN) is neutral in the relevant sense, since it is silent towards ultimate questions. This paper examines whether this is so, especially when discussing the content of the science curriculum in public education. In this context, advocates of the status of Intelligent Design creationism as a scientific theory argue that MN arbitrarily dismisses a type of supernatural agency that is fundamental to several branches of theistic belief. Drawing on Thomas Nagel’s position, the paper contends that MN assumes a meta-scientific position to which either god does not exist, or god does not intervene in cosmic history. To that extend, MN cannot be reconciled with LPRN. However, this conclusion does not entail that hypotheses that suggest supernatural agency must be included in the science curriculum. Creationist theories may remain excluded to the extent that they fall short of the standards that are required for a proficient scientific account. But, crucially, it will not be their religious character what keeps them out of the classroom, but a series of pseudoscientific wrongs which are non-exclusive to supernaturalistic hypotheses. The paper thus suggests the exploration of an alternative demarcation strategy, one that is consistent with LPRN.