Abstract: Objective: The objective of this study was to compare energy values of wheat or barley derived from the direct and regression methods for broiler chickens. Materials and Methods: Seven diets consisted of a corn-soybean meal-based basal diet, 2 semi-purified diets mixed to contain wheat or barley as the sole source of energy and 4 test diets prepared by supplementing the basal diet with wheat or barley at 15 or 30%, respectively. Chromic oxide was used as an indigestible index to calculate the metabolizability. A total of 504 21-days-old male Ross 308 broiler chickens were allocated based on body weight into 7 treatments with 6 replicate cages and 12 birds per cage by using a randomized complete block design. From d 21, birds were fed the experimental diets for 5-d and excreta samples were collected from d 24-26. Results: The apparent metabolizable energy (AME) of the wheat using the direct method (2,934 kcal kg–1 DM) was not different from the value (3,026 kcal kg–1 DM) derived from the regression method. However, the AME of barley obtained from the direct method (2,730 kcal kg–1 DM) differed from the value (2,970 kcal kg–1 DM) measured by the regression method. Conclusion: The direct and regression methods gave different estimates of the AME in barley but not in wheat fed to broiler chickens.