Periplaneta americana residue is a byproduct of using Periplaneta americana in pharmaceutical research and development for extracting active ingredients. Three hundred Three-yellow chickens were selected for the experiment and randomly divided into 6 groups (5 replications per group, 10 chickens per replicate): the control group (group A) was fed a basal ration, and the experimental groups (groups B, C, D, E, and F) were fed experimental diets in which P. americana residue replaced puffed soybean meal at proportions of 20, 40, 60, 80, and 100%, respectively, for a period of 42 d. The aim was to assess the impact of different levels of P. americana residue on the growth, survival, intestinal morphology, digestive enzyme activity, intestinal flora, and intestinal transcriptional responses of Three-yellow chickens. The results indicated that the increase in P. americana residue levels had a linear and quadratic impact on the average daily gain (ADG) and feed conversion ratio (FCR), respectively. The ADG was notably greater in the 40% group than in the 100% group, while the FCR was significantly lower in the 20% and 40% groups than in the 100% group (P < 0.05). Protease, lipase, and amylase activities exhibited a quadratic increase with increasing concentrations of P. americana residue (P < 0.05). Protease and lipase activities were notably greater in the 20% and 40% groups than in the 0% group (control group), amylase activity was significantly greater in the 40% group than in the 0% group (control group) (P < 0.05). Duodenal crypt depth (CD) decreased quadratically with increasing P. americana residue (P < 0.05). The duodenal villus height/crypt depth ratio (V/C) was significantly lower in the 100% group than in the 60% group (P < 0.05). The intestinal villus height (VH) increased quadratically with increasing levels of P. americana residue. The VH in the 60% group was significantly greater than that in the 0% (control group), 20, 80, and 100% groups (P < 0.05). The Chao and Ace indices demonstrated linear and quadratic increases with increasing levels of P. americana residue, while the Pd index showed a quadratic increase with increasing levels of P. americana residue (P < 0.05). The relative abundance profile of Lactobacillus exhibited a linear and quadratic decrease with increasing levels of P. americana residue, with the 100% group showing a significantly lower abundance than the 0% (control group) and 40% groups (P < 0.05). The transcriptome results showed that P. americana residue could enhance the digestive system by promoting vitamin, fat, carbohydrate digestion and absorption, cholesterol metabolism, etc. In conclusion, P. americana residue can replace puffed soybean meal without negatively affecting the growth performance of three-yellow chickens. The low and medium groups had positive effects on the growth performance, digestive enzyme activity, intestinal morphology, intestinal flora, and substance digestion and absorption of three-yellow chickens. The recommended replacement of P. americana residue for puffed soybean meal in the diets of three-yellow chickens ranged from 20% to 60%.