The aim of this experiment was to research and validate the prediction model of protein requirement of Guangming No.2 broiler. First, the protein growth requirement of broiler is measured (growth test), and then combined with the protein maintenance requirement measured in the previous stage of this project by the writer's team, a prediction model of protein requirement of broiler is established and verified (validation test). The growth test was designed with a completely randomized trial, and 132 1-day-old Guangming No.2 male and female broilers with similar body weight were selected, and they were divided into 2 treatment groups according to sex, with 6 replicates and 22 birds per replicate. They were fed a corn-soybean meal diet until 6 weeks of age. The body weight and body protein weight of broilers were measured weekly, and a linear regression equation between body weight and body protein weight was established, and the slope was the growth requirement of broiler protein. The validation test was carried out in three phases (0-10 days of age, 11-24 days of age, and 25-39 days of age). At 0, 11 and 25 days of age, 200, 150 and 100 healthy Guangming No.2 broilers were selected, and the broilers were randomly divided into control group and experimental group, with 5 replicates in each treatment group. The control and experimental diets were formulated according to Aviagen recommended standards and predicted protein requirements, respectively. The growth test showed that the production performance of 21 d and 42 d Guangming No.2 male broilers was significantly higher than that of female broilers (P < 0.01). The linear regression equation of broiler body weight and body protein weight was established in stages, and the regression model all reached significant level (P < 0.001), with R2 of 0.99 in the early stage and 0.98 in the late stage. According to the linear regression equation, the growth requirement of body protein is calculated, and the growth requirement of diet protein is further calculated according to the deposition efficiency of dietary protein, which was every gram of ADG requires 0.251 and 0.228 g diet protein (the quantity unit was expressed in g·g-1 ADG) for male and female broilers aged 1-21days, and 0.272 and 0.268 g·g-1 ADG for male and female broilers aged 22-42days, respectively. Combined with the protein maintenance requirements measured in the previous stage of the writer's team, a prediction model of protein requirements was established: 0-21 days old: CPR=2.365×BW0.75 +0.251×ADG (male), CPR=3.165×BW0.75 +0.228×ADG (female); 22-42 days of age: CPR=2.955×BW0.75 +0.272×ADG (male), CPR=2.560×BW0.75 +0.268×ADG (female). The verification test showed that the body weight and ADG of broilers in experimental group were higher than those in control group at 0-10 days of age (P < 0.05), and lower than those in control group at 11-24 days of age and 25-39 days of age (P < 0.05). The protein content of the diet in the experimental group was predicted by the model based on the body weight, daily gain and feed intake of AA broilers. Due to the differences in the growth curves and feed intake changes between Guangming No.2 broilers and AA broilers, the prediction results may be inaccurate, which does not indicate that there are problems in the prediction model. Both the control group and the experimental group used Guangming No.2 broilers, so the protein requirement of the broiler diet could be calculated based on the actual body weight and daily gain, using a prediction model. After comparison, it was found that only the dietary protein supply of broilers in experimental group aged 1-21 days exceeded the requirement by 12.2%, the rest of the groups did not exceed 10%. These results indicate that the prediction model of dietary protein requirement of Guangming No.2 broiler proposed in this study is basically accurate.