The pelletizing of fine-grained mineral raw materials for ferrous metallurgy is studied. Relationships between the compressibility of metallurgical charges and the nature of the raw materials and process factors of the pelletizing process were studied in laboratory conditions. A new measure, such as the compressibility of ground materials, was introduced to quantify and compare compression performances. A mathematical description of the relationships between the compression coefficient and the following factors was formulated through nonlinear regression analysis and experiment design theory methods: the plasticizer content (varying from 0 to 50%), moisture content of the charge (from 0 to 10%), hardness of the particles according to the mineralogical scale (from 2 to 6 units), dynamic viscosity of the binder (from 1 to 657 mPa · sec), amount of the carbon-containing component (from 10 to 90%), and particle sizes of the iron- and carbon-containing components in the charge (from 1 to 4 mm). The compaction pressure range was limited to 220 MPa. Three mathematical models were developed to establish relationships between the charge compressibility and the specified factors as polynomial dependencies and as a Lorentz function. The quality of the models was verified using standard statistical indicators, including the Cochran and Fisher tests and the average relative error. Analysis of the models involved the solution of relevant optimization problems. Extremes of the functions for metallurgical charges were identified and process recommendations were made. The results have practical implications for improving the compressibility of charges from mineral raw materials and developing optimal pelletizing methods.