The most recent sapropel (S1) in the eastern Mediterranean has been extensively investigated with geochemical and rock-magnetic techniques. Different redox conditions prevailed in different zones of the sediment through time. The oxidized sapropel zone is particularly interesting, because earlier studies indicated that new magnetic material was formed, including possible magnetosomes. Here we utilize component analyses of isothermal remanent magnetization (IRM) acquisition curves and the analysis of first-order reversal curves (FORC) to further investigate the magnetic mineralogy. In the entire box core ABC26, the original input of eolian dust consisted of both magnetite and hematite. In the oxidized sapropel and in the active oxidation zone, an additional magnetite component is present. This magnetite component has a higher coercivity than the eolian magnetite and a very small coercivity dispersion, suggesting a narrow grain size distribution. This is a strong indication that magnetosomes, which are formed in the active oxidation zone, are the magnetic carriers of this coercivity fraction. FORC diagrams support these findings. The S ratio is forwardly modeled for mixed magnetic mineralogies with varying coercivity distributions to explain the down core S ratio behavior: close to 1 in the top sediment and in the oxidized sapropel, a drop in values in the active oxidation zone, and again close to 1 in the (syn)sapropel. The observed S ratio pattern in the oxidized sapropel and in the active oxidation zone can be explained by the recovered IRM components. Whereas both oxidized sapropel and active oxidation zone contain the extra magnetite component, their S ratios are different because of the differences in coercivity characteristics (coercivity and dispersion) of the two magnetites in these zones. Thus, in these zones the S ratio does not reflect variations in the relative contributions of hematite to magnetite but variations in the characteristics of the individual magnetite assemblages. Also in the rest of the core, the S ratio depends on the coercivity characteristics of the magnetite component rather than on the relative contributions of high- versus low-coercivity minerals. The S ratio thus appears to be an unsuitable parameter to describe variations in the magnetic mineralogy, especially when more than two components are present. Therefore the classical interpretation of the S ratio should be treated with caution.