Abstract: A modified true metabolizable energy bioassay was used to determine the nitrogen-corrected apparent (AMEn) and true (TMEn) metabolizable energy of various feed ingredients for ducks. In each of two experiments, which lasted 102 h with a 54-h excreta collection period following a 48-h period of feed withdrawal, six male White Pekin ducks were assigned to each of three test feed ingredients and dextrose. Dextrose-fed ducks were used for estimation of endogenous losses of energy and nitrogen. The test ingredients consisted of corn, bakery byproduct and wheat red dog in experiment 1 and corn, corn gluten meal and soybean meal in experiment 2. The AMEn and TMEn values of corn were similar in the first and second experiments at 3.322 and 3.358 and 3.289 and 3.528 kcal/g, respectively. The AMEn and TMEn values of corn were higher (P < 0.05) than those of wheat red dog (2.519 and 2.662 kcal/g), but lower (P < 0.05) than those of bakery byproduct (3.75 and 3.896 kcal/g). Corn gluten meal had the highest (P < 0.05) AMEn and TMEn values (3.695 and 3.934 kcal/g) among three feed ingredients. The AMEn and TMEn value of soybean meal (2.562 and 2.799 kcal/g) were lower (P < 0.05) than those of corn. The study provide new information on AMEn and TMEn values for bakery byproduct, corn, corn gluten meal, soybean meal and wheat red dog; and demonstrated that the energy values, for White Pekin ducks, of bakery byproduct and corn gluten meal are greater than those of corn or soybean meal.