Combining the abundance and inexpensiveness of their constituent elements with their atomic dispersion, atomically dispersed Fe–N–C catalysts represent the most promising alternative to precious-metal-based materials in proton exchange membrane (PEM) fuel cells. Due to the high temperatures involved in their synthesis and the sensitivity of Fe ions toward carbothermal reduction, current synthetic methods are intrinsically limited in type and amount of the desired, catalytically active Fe–N4 sites, and high active site densities have been out of reach (dilemma of Fe–N–C catalysts). We herein identify a paradigm change in the synthesis of Fe–N–C catalysts arising from the developments of other M–N–C single-atom catalysts. Supported by DFT calculations we propose fundamental principles for the synthesis of M–N–C materials. We further exploit the proposed principles in a novel synthetic strategy to surpass the dilemma of Fe–N–C catalysts. The selective formation of tetrapyrrolic Zn–N4 sites in a tailor-made Zn–N–C material is utilized as an active-site imprint for the preparation of a corresponding Fe–N–C catalyst. By successive low- and high-temperature ion exchange reactions, we obtain a phase-pure Fe–N–C catalyst, with a high loading of atomically dispersed Fe (>3 wt %). Moreover, the catalyst is entirely composed of tetrapyrrolic Fe–N4 sites. The density of tetrapyrrolic Fe–N4 sites is more than six times as high as for previously reported tetrapyrrolic single-site Fe–N–C fuel cell catalysts.