Apparent digestibility of dry or moist cat food was determined by both total feces collection method and the indicator method using Cr₂O₃ as an indicator. In Experiment 1, adult cats (n=8) were given free access to a canned food consisted mainly of fish meat (71.1% moisture and 21.5%crude protein). Digestibility of dry matter, crude protein, crude fat (ether extract), and N-free extract was determined by the two methods during week 1, week 2, and week 3 after a 5-d precollection period. Digestibility of these components except for N-free extract was consistent among the three time periods independent of the determination methods. In addition, fecal dry matter excretion was almost proportional to dry matter intake throughout the 3-wk period. In Experiment 2, adult cats (n = 6) were given free access to an extruded dry diet containing a plenty of vegetable ingredients (6.1% moisture and 27.3% crude protein). When digestibility was determined by the indicator method, digestibility was consistent among the time periods of week 1, week 2, and week 3 after a 5-d precollection period. In contrast, when digestibility was determined by the total feces collection method, the value tended to vary with time periods except for crude fat (acid-ether extract), and the tendency was most obvious with crude protein. Fecal dry matter excretion did not parallel to dry mater intake in Experiment 2. These results suggested that the indicator method was preferable to the total feces collection method in determining apparent digestibility of extruded dry cat foods, probably because of uneven excretion of fecal dry matter. However, a 5-d precollection period seemed sufficient to obtain accurate digestibility even in carnivorous cats, at least when the indicator method is adopted.