The study is focused on the problem of using geophysical data to estimate brittleness of rock masses for the needs of petroleum industry. Three main developed ways to estimate brittleness—mineral-based, log-based, and elastic-based brittleness indices—are discussed from the perspective of scaling factor. The study highlights the contradictions between brittleness indices calculated from the same data using various ways of introducing brittleness. These contradictions are explained by scaling factor, as geophysical data used for brittleness estimation are typically obtained at different spatial and temporal scales. A model based on the effective medium theory is used to understand the relationships between inner structure of inhomogeneous rocks and their brittleness indices estimated from laboratory tests on core samples as well as log data analysis.