Mercury porosimetry was used to study the pore‐size distribution of both green and partially sintered alumina compacts. Experiments showed that using the mercury‐penetration volume data to study the pore structure changes during sintering for a compact with a bimodal pore‐size distribution could lead to wrong conclusions because the mercury‐penetration volume was not linearly proportional to the porosity. Instead, the porosity distribution should be used. When the porosity of large pores was<10%, some large pores within the compact were observed to become isolated or channeled to the compact surface; this caused errors in the porosity distribution measured by mercury p