Several kinds of tests exist to characterize the galling of metals, such as that specified in ASTM Standard G98. While the testing procedure is accurate and robust, the analysis of the specimenʼs surfaces (${\rm area}=1.2\;{\rm cm}$) for the determination of the critical pressure of galling remains subject to operator judgment. Based on the surfaceʼs topography analyses, we propose a methodology to express the probability of galling according to the macroscopic pressure load. After performing galling tests on 304L stainless steel, a two-step segmentation of the Sq parameter (root mean square of surface amplitude) computed from local roughness maps ($100\;\mu {\rm m}\times 100\;\mu {\rm m}$) enables us to distinguish two tribological processes. The first step represents the abrasive wear (erosion) and the second one the adhesive wear (galling). The total areas of both regions are highly relevant to quantify galling and erosion processes. Then, a one-parameter phenomenological model is proposed to objectively determine the evolution of non-galled relative area Ae versus the pressure load P, with high accuracy (${{A}_{e}}=100/(1+a{{P}^{2}}$) with $a={{0.54}_{\pm 0.07}}\times {{10}^{-3}}\;M{\rm \ }P{{a}^{-2}}$ and with ${{R}^{2}}=0.98$). From this model, the critical pressure of galling is found to be equal to $43{\rm MPa}$. The ${{S}_{5\,V}}$ roughness parameter (the five deepest valleys in the galled regionʼs surface) is the most relevant roughness parameter for the quantification of damages in the 'galling region'. The significant valleys' depths increase from 10 μm–250 μm when the pressure increases from 11–350 MPa, according to a power law (${{S}_{5\,V}}=4.2{{P}^{0.75}}$, with ${{R}^{2}}=0.93$).