Hematite and goethite are widely occurring chromogenic iron oxides in soils and sediments that are sensitive to climatic dry/wet shifts. However, only by accurately quantifying the content or ratio of hematite and goethite can they be applied reliably to palaeoclimate reconstruction. Compared to the Loess Plateau of China, hematite in the soils of southern China has not been sufficiently studied. We used diffuse reflectance spectroscopy (abbreviation DRS, including the first-derivative curves and the second-derivative curves of the Kubelka–Munk remission functions), combined with ignition at 950 °C, and X-ray fluorescence (XRF) to quantify the hematite content of four tropical-margin iron-rich soil profiles with different matrix compositions in the Leizhou Peninsula, China. We also examined the application of hematite quantification parameters in soils with different matrix compositions under the same climatic conditions. Our main findings are as follows: (i) DRS first-derivative curves can reflect the presence of goethite and hematite in soils, and their relative contents can be compared within the same profile. (ii) The second-derivative curve of the Kubelka–Munk remission functions can reflect the relative proportions of goethite and hematite and provide information about the degree of Al substitution. (iii) Combined with calibration equations, soil redness can reliably quantify the hematite content, but it is necessary to consider the effect of mucilage envelopes in the process of hematite formation. Additionally, we summarize various methods used for quantifying hematite, and the influence of soil matrix compositions, with the aim of providing a reference for hematite quantification elsewhere. We also propose a new indicator (ΔHmRed/HmRed) to help detect iron hydroxide/iron oxide changes in soils.