Feed efficiency (FE) is an increasingly important trait for a sustainable dairy sector with variation among individuals. Several definitions have been proposed to characterize FE. This study aimed to evaluate if the hierarchy between animals for FE was repeatable at different physiological stages and if different metrics gave similar results for characterizing this FE. Thirty-three lactation goats (12 Alpine and 21 Saanen) born the same year were housed in individual pens at three periods (middle (P1) and end (P2) of first lactation and middle of second lactation (P3)). They were fed a total mixed ration (TMR) ad libitum. The TMR consisted of 20% meadow hay, 30% chopped dried alfalfa, 30% pressed sugar beet pulp silage and 20% concentrate on a DM basis. FE was characterized by residuals of equations obtained with four metrics: 1/ dry matter intake (DMI, kg/day) was expressed on bodyweight (BW, kg), Fat and Protein Corrected Milk Yield (FPCM, kg/day) and stage of lactation (days in milk) or RFI (Residual Feed Intake), 2/ Ratio of FPCM/DMI in function of FPCM/BW or RMI, 3/ observed FPCM on estimated FCPM by dietary allowance and 4/ ratio of FPCM on Net Energy intake. For the metrics 3 and 4, the INRA Feeding System for Ruminants was used (INRA 2018). DMI was higher in P4 (3.00 ± 0.282) than in P2 (2.63 ± 0.244) and in P3 (2.49 ± 0.216). BW was lower in P2 (49.2 ± 4.03) than in P3 (59.8 ± 5.69) and P4 (59.3 ± 4.77). FPCM differed between periods (P1: 3.43 ± 0.474; P2: 2.07 ± 0.448; P3:3.99 ± 0.574). The stage of lactation was not significant for the first metric as it was highly correlated to FCPM (r =
0.85, n = 33). For the same reason, the period effect was not significant in the first residual. The period effect was significant for the three-last metrics, with a lower efficiency for the second period (end of lactation). The goat effect was always significant and the coefficients of correlation between the four metrics were always significant. This means that there was a stable hierarchy between animals among the three periods meaning that feed efficiency might be a repeatable trait regardless of the metric used and that the four metrics gave similar results to characterize the animals. The between goats’ differences in feed efficiency might be explained by differences in digestive efficiency (influence of retention time and/or microbiota composition in the rumen) and/or in metabolic efficiency (role of body reserves, protein turnover, heat production, . . .). As these results were obtained in indoors goats fed a balanced TMR diet, they need to be tested in other dietary contexts to check if FE is still a resilient trait.