Several feed efficiency (FE) metrics are currently used in livestock production to select for improved FE. Whether or not different FE metrics similarly estimate physiological characteristics in chickens of diverging FE has not been reported so far. This study aimed to assess potential differences in feed intake (FI), performance, and nutrient excretion in broiler chickens of diverging FE when ranked according to their residual FI (RFI), residual BW gain (RBG), RFI and BW gain (RIG), and G:F between d 7 and 35 of life. The FI was determined daily and BW was recorded once a week. The ranking of chickens into good, medium, and poor FE groups was completed separately for each FE metric. Freshly dropped excreta were collected for pH and DM measurements on d 30 to 32 of life and total excreta for determination of nutrient excretion was collected on d 34 to 36 of life. Relationships among FE metrics were evaluated using regression analysis showing that RFI, RIG, and G:F were more related to each other than to RBG. The FE values greatly varied among chickens for all FE metrics and chickens did not always cluster within the same FE group when using RFI, RIG, RBG, and G:F as the FE metrics because of the calculation approaches. Due to sex-related differences in performance, data of male and female chickens were analyzed separately. The RFI and RIG metrics showed a linear increase (P < 0.01) in total FI from good to poor FE in male and female chickens, whereas G:F showed this effect (P <= 0.011) only when BW gain was standardized to 1,500 g. The RBG did not clearly select chickens of enhanced total BW gain and only tended (P < 0.1) to select for greater BW gain from good to poor FE in female chickens. Excreta pH linearly decreased by 0.7 log units and DM content increased in males from good to poor FE when using RFI and RIG, respectively (P < 0.01). In both sexes, RFI (P < 0.05) and RIG metrics (P <= 0.06) showed a linear increase in daily nitrogen excretion from good to poor FE. In conclusion, results demonstrate that selection of the metric used to determine the FE of chickens modified the results obtained for comparison of production parameters and nutrient excretion among FE groups. Thereby, the RFI, RIG, and G:F metrics were beneficial in selecting the most feed efficient chickens to reduce feed costs, whereas the use of RFI and RIG may be better to select chickens with improved nitrogen retention and thus reduced excretion of an environmental pollutant.