The methods for determining the pore volume distribution by the size of porous materials are analyzed and it is shown that the most rigorous are methods in which capillary effects do not appear. It is shown that mercury porosimetry and the method of liquid displacement by gas have a large error in the region of large pores. This is due to the percolation effect. It is proved that determining the distribution of pore volume from measurements of air flow gives too distorted results, so its use is not advisable.