Pellet-group counts were made with penned mule deer (Odocoileus hemionus) over a period of 12 years. The defecation rates observed were somewhat higher than those commonly assumed, being from 13 to 14 groups per day. Weights of pellet groups, the number of pellets per group, and pellet size were too variable to provide any assistance in identifying fecal deposits of individual animals. Foraging juvenile animals exhibited higher rates of defecation than did mature animals; therefore, age classifications are important to accurate censuses. The rates of yearlings seldom differed significantly from older animals. The level of forage intake affected the rate of defecation. Rates must be expressed in terms of the weight of the animal; otherwise, juvenile animals with low forage consumption and high rates mask the effect. Also, the kind of forage eaten affected the number of defecations. Barring mildly laxative effects of concentrates and possibly succulent forage, coarse woody material usually produced higher defecation rates than the same amounts of more nutritious food. Data obtained from field counts with known populations of deer gave generally lower values than were found in the pens, but failure to locate pellets may have been an important factor. The validity of counts in dense vegetation in the summer is highly questionable. Statistical analyses of counts made on winter ranges indicate more than 200 plots of 100 square feet each in size may be required for reasonable accuracy with high deer populations. The 100-square-foot plot size, though a convenient area for counting, is possibly too small a sample. Four such plots in a cluster seem to be a better basis for the individual observation. Further modification of the plot to 108.9 square feet so that each observation would represent 0.01 of an acre would facilitate computation. Counting pellet groups is a technique used to determine populations and distribution of game animals. This technique necessitates knowing the defecation rate of the animal to establish the size of the population, otherwise pellet-group counts only indicate relative numbers and distribution of animals. Furthermore, where animals occupy a range for only part of the year, uncertainty with respect to arrival and departure dates of the animal may prevent the determination of animal numbers. Bennett et al. (1940) first proposed the pellet-group technique for comparing white-tailed deer (Odocoileus virginianus) populations, but they provided no data on defecation rates. Rasmussen and Doman (1943) made counts of pellet groups in enclosures where the mule deer populations were known. Their findings first made possible the conversion of pelletgroup counts to deer populations. More recently data have been reported that are somewhat at variance with figures rereported by them. The data reported herein provide additional information on defecation rates of mule deer and some problems encountered in sampling and identifying pellet groups in the field.