Initial published work suggested that Fe isotope fractionations recorded in sediments were a product of biological activity. Experiments and measurements of natural samples now indicate that Fe isotope fractionation can be the product of both biological and inorganic processes. Sedimentary iron sulphides provide unique information about the evolution of early life which developed under anoxic conditions. It is in these sedimentary Fe-S species and in particular in Archean and Proterozoic pyrites that the largest Fe isotope variations (up to a range of ~5‰ for δ56/54Fe) have been measured. Most research has focussed on potential processes responsible for the formation of a 56Fe depleted Fe(II) pool from which iron sulphides would precipitate without additional fractionation, recording the light Fe isotope composition of the pool. Much less attention has been given to the possibility that the iron sulphide forming mechanisms themselves could produce significant fractionations. The Fe-S system constitutes a diverse group of stable and metastable phases, the ultimate Fe sequestrating phase being pyrite. The aim of this study was to examine experimentally where Fe isotope fractionations occur during the abiotic formation of iron sulphides in order to assess whether or not the measured Fe isotope signatures in natural pyrite could be explained by chemical mechanisms only. Both analytical and experimental protocols were developed in order to determine the partition of Fe isotopes for each step towards diagenetic pyrite formation. 56/54Fe and 57/54Fe ratios were measured on an IsoProbe-P Micromass MC-ICP-MS, and all experiments were performed under oxygen-free N2 atmosphere. Supporting previously published data, the results indicate that the precipitation of the nanoparticulate iron(II) monosulphide mackinawite (FeSm) kinetically fractionates lighter isotopes with initial fractionations of Δ56FeFe(II)aq-FeS = 1.17 ± 0.16 ‰ at 25°C and Δ56FeFe(II)aq-FeS = 0.98 ± 0.16 ‰ at 2°C. The rate of isotopic exchange between Fe(II)aq and FeSm decreases as FeSm nanoparticles grow. Fe isotope exchange kinetics are consistent with i) FeSm nanoparticles that have a core–shell structure, in which case Fe isotope mobility is restricted to exchange between the surface shell and the solution and ii) a nanoparticle growth via an aggregation– growth mechanism. Because of the structure of FeSm nanoparticles, the approach to isotopic equilibrium is kinetically restricted at low temperatures. The equilibrium Fe isotope fractionation between Fe2+ aq and FeSm was determined using the three isotope method and is Δ56FeFe(II)-FeS = -0.33 ± 0.12 ‰ at 25°C and Δ56FeFe(II)-FeS = -0.52 ± 0.16 ‰ at 2°C. This suggests that at equilibrium, FeSm incorporates heavier isotopes with respect to Fe2+ aq, and the isotopic composition of most naturally occurring FeSm does not represent equilibrium. During pyrite formation, pyrite incorporates kinetically lighter isotopes with a fractionation Δ56FeFeS-pyrite ~ 2.2 ‰. Because pyrite is sparingly soluble in sedimentary environments, isotope exchange is prevented and pyrite does not equilibrate with its Fe(II) source. Combined fractionation factors between Fe2+ aq, mackinawite (FeSm) and pyrite permit the generation of pyrite with Fe isotope signatures that encapsulate the full range of sedimentary δ56Fepyrite recorded in both Archean and modern sediments. Archean Fe isotope excursions reflect various degrees of pyritisation, extent of Fe(II)aq utilisation, and variations in source composition rather than microbial dissimilatory Fe(III) reduction only. Our results show that sedimentary pyrite is not a passive recorder of the Fe isotope composition of the reactive Fe(II) reservoir forming pyrite. It is the formation process itself that influences pyrite Fe isotope signatures with consequent implications for the interpretation of sedimentary pyrite Fe isotope compositions throughout geological time.