This paper examined the effect of N-correction on the true metabolizable energy (TME) values of mixed diets and single ingredients using mature White Leghorn roosters. The N-corrected TME (TMEn) values were consistently higher than the corresponding TME values, the results being significant (P less than .05) in about one-half of the samples tested. This indicated a net loss in body N in the birds during the experiments. The correlations between N input and N correction were highly (P less than .01) significant. This, however, explained only 46 to 48% of the variation in the TMEn data as being due to differences in the amount of N input. In view of the present results and the abnormal condition (fasted birds) in the TME assay, it is doubtful that birds would retain enough N from a single feeding to revert to a positive body N equilibrium. Therefore, positive N retention in such an assay would be indicative of incomplete feed passage through the birds. This appears to be the only valid reason for applying N correction to TME values. However, the variations associated with N correction of TME data raise some doubt concerning the validity of this assay as compared with the conventional apparent metabolizable energy assay.