Abstract: Abstract: The pollution production coefficient of sheep-breeding is lacking in recent years, since most previous studies focused mainly on pigs, cattle, and chickens in the livestock and poultry breeding industries. This study aims to systematically explore the producing characteristics of pollutants from large-scale sheep farms in Northern China, and thereby to construct the pollution production coefficient for the sheep breeding industry. Three typical sheep farms were selected for sentinel monitoring experiments in Northeast, North and Northwest China. The sheep were divided into four groups according to the feeding stages, including the nursling, fattening, mature, and barren stages. In each group, five normal, healthy and similar weight sheep were randomly selected using the age and gender of different sheep breeds. A metabolic cage was adopted in the experiment, where feed and water were supplied regularly and quantitatively. The trial period was from July to September 2019. The monitoring items included the daily feed intake, water intake, as well as the amount of feces and urine of each sheep. The feed of each group, the feces and urine samples of each sheep were collected for 3 consecutive days to determine physical and chemical indicators for the pollution production coefficients. The results showed that there was great difference in the regularity of pollutant generation for the breed, age, and gender in a sheep farm. The daily feed and water intake of Liaoning cashmere goat in northeast China, the crossbred F1 of Dorper and Hu sheep in north China, and the Tan sheep in northwest China were 1.45, 1.29, and 2.24 kg, while, 1.82、2.73, and 0.72 L, respectively. The daily production of feces and urine for the Liaoning cashmere goat, the crossbred F1 sheep, and Tan sheep were 0.69, 0.61, and 0.94 kg, 0.77, 1.19, and 0.42 L, respectively. The daily fresh weight of feces and fresh weight of feed, dry weight of feces and dry weight of feed, urine volume and water consumption all showed significant positive correlation (0.8