A direct measuring method for the liquid penetration rate into tablets was devised and compared with the usual powder method. The effect of the diameter and shape of tablets on the penetration rate was examined by using a two-dimensional model and confirmed to be negligible in practice. Directly compressed tablets of several powders used as excipients were tested as samples. The weight variation of tablets had little effect on penetration. Since the penetration rate depends on the compression pressure, all tablets were prepared at a constant pressure of 215 kg/cm2. Both in the present method and in the usual powder method most samples were in conformity with Washburn's equation, but the rate constants (K) were considerably different in the two methods. The results for crystalline cellulose could be described by such an equation in the case of the powder method but not in the case of the tablet method. The reason is considered to be the swelling of tablets and the accompanying generation of cracks. In this case, an L vs. t plot showed good linearity instead of an L2 vs. t plot. The penetration patterns of crystalline cellulose tablets were also different from those of the other samples. Most of the tablets were not disintegrated after completion of penetration ; calcium carbonate, lactose and crystalline cellulose were exceptions.