AbstractNitrogen is the most abundant element in the Earth's atmosphere, yet its geochemical behavior and distribution among the various reservoirs (atmosphere, crust, mantle, and core) remain poorly understood. Although estimates of N and C fluxes in the mantle vary, there is a consensus regarding the disparity between input and output, leading to an increase in N and C contents in the mantle. The low solubility of N in mantle minerals raises questions about possible N or C storage in the mantle. Evidence suggests that Fe–N–C phases, such as Fe₃C, Fe₇C₃, ε-Fe₃N, metals, and non-stoichiometric carbonitrides, may be accessory phases at mantle pressure and temperature conditions, and thus potential hosts of C and N in the deep mantle. To investigate the phase relations and melting behavior in the (Fe,Ni)–N–C system, 19 experiments were conducted with varying starting compositions at 10 GPa and 1000–1400 °C. The results indicate that carbides, nitrides, carbonitrides, nitrocarbides, Fe(Ni)-metal, Fe-oxides, and diamond are stable at deep upper mantle pressure conditions. However, the compositions of naturally occurring nitrocarbides with high C and N contents, as found in diamond inclusions, could not be reproduced in the experiments. The significant incorporation of Ni in the experimental phases, which is also not observed in natural carbonitrides and nitrocarbides, suggests their formation in Ni-poor regimes. The solidus temperatures of the N- and C-rich systems are well below the adiabatic temperatures of the surrounding mantle. Therefore, it is hypothesized that cold regions in subduction zones, such as within or at the edge of a C- and N-rich subducted plate, are the likely formation environment for solid Fe–C–N phases.