<p>Among the various stances of scientific realism, experimental realism is unique in prioritizing successful intervention over representation as the privileged route to acquire warranted knowledge of reality. Experimental realism advocates for the belief in the entities amenable to manipulation while maintaining skepticism toward the theories that attempt to describe them. In its original formulation, the continued manipulation of a putative entity is posited as the sole criterion for its reality. This criterion has recently been complemented by additional ontic properties — such as observability, countability, and persistence across sciences — that a putative entity ought to satisfy by virtue of its existence. Here, drawing from the writings of eighteenth-century chemists, I examine the reliability of experimental realism in light of the phlogiston theory. I argue that phlogiston, despite its non-existence, qualifies as real due to its ability to meet the ontic properties, thereby challenging the tenability of experimental realism. This paradoxical conclusion suggests that intervention, without a comprehensive representation, is an erring guide to reality and may endow metaphysical status to unreal entities that are loosely characterized. Overall, this essay casts doubt on the general applicability of the criteria to infer the existence of putative entities outlined by experimental realism.</p>