[en] When submitted to mercury porosimetry, some materials are penetrated by mercury whereas others, among the most porous, are densified by the isostatic pressure. Notably, this is the case for materials whose structure is made of particles aggregated into filament-like clusters that are interconnected in a 3-D array. Indeed, that kind of material undergoes a volume variation due to hierarchical pore collapse. In the case of intrusion, the mercury porosimetry data are classically analyzed by the Washburn equation. In the case of hierarchical pore collapse, data can be correctly analyzed by the collapse model equation. Using an equation that does not correspond to the mechanism leads to large errors in the pore size distribution. Thus, an accurate data analysis requires prior determination of the mechanism leading to the volume variation recorded as a function of the pressure. The present work particularly examines the complex and unusual behavior of partially pyrolyzed resorcinol-formaldehyde gels when submitted to mercury porosimetry. The unusual behavior encountered complicates the mechanism identification and, therefore, the equation selection. However, the major part of the volume distribution as a function of the pore size can be determined with a good accuracy.