The 2.4Ga Hotazel Formation is a cyclically interlayered sequence of banded iron formation (BIF) and manganese-rich sedimentary rock at the uppermost part of the Neoarchaean-Palaeoproterozoic Transvaal Supergroup in South Africa. It represents an unusual stratigraphic association in the context of the origin of BIF and the coevolution of oxygen and life on early Earth and hence bears special relevance to the environmental conditions and processes that characterized the period leading up to the Great Oxidation Event (GOE) at ca. 2.3Ga. The mineral assemblages that characterize the Hotazel rocks are dominated by carbonate, silicate and oxide minerals, which are traditionally interpreted as predominantly diagenetic in origin, particularly the carbonates. By contrast, primary mineral assemblages are inferred to have been dominated by ferric oxyhydroxides and tetravalent manganese oxides, which show no preservation in the rock record and consequently hinder reconstruction of environmental conditions during sedimentation. Here, we revisit the Hotazel succession with a focus on its bulk-rock and carbonate-specific mineralogical, geochemical and stable isotope (C, Fe) composition by applying for the first time a high-resolution stratigraphic approach to sampling and analysis. Our main aim is to constrain the precursor mineralogy to the Fe- and Mn-rich facies in the Hotazel strata in order to unravel the redox conditions behind the massive cyclic deposition of Fe and Mn at the onset of the GOE. Our carbonate-specific results question traditional diagenetic models for the development of the carbonate fraction of the rocks and instead place the origin of much of the present mineralogy on water-column processes in a stratified basin characterized by successive redox pathways with changing water depth. These pathways exploited a series of thermodynamically predictable electron acceptors for organic carbon recycling, which included – probably for the first time in Earth history – aqueous Mn(III) and O2 as electron acceptors for the oxidation of both Fe(II) and organic carbon. The emergence of Mn(III) was also critical for the development of a Mn redox shuttle, which led to effective water-column stratification between aqueous Mn and Fe in the depositional basin. We conclude that the first known record of Mn(II) to Mn(III) oxidation as recorded in the Hotazel Formation must be a fundamentally diagnostic step in the redox evolution of the oceans and atmosphere in the lead-up to the GOE.