Most prediction equations of feedlot performance cover average daily gains (ADG) below 1.5 kg/d. The question is whether these equations would be accurate for higher ADG. Medium-to-large-framed weanling and yearling steer calves were fed individually using a transponder-electronic gate system in three feedlot trials and fed in groups in two trials. Average daily gain ranged from .9 to 2.1 kg. Steers were implanted, fed to the same degree of finish, and slaughtered to obtain dressing percentage and cold carcass weight. Initial carcass weights were estimated from initial slaughter groups. Dietary ME concentration (i.e., 10.5, 11.5, 12.0, and 12.5 MJ ME/kg of DM, where 1 Mcal = 4.184 MJ), initial weight, dry matter intake (DMI), and days fed could explain 57% of the variation in ADG. The partial contribution of DMI to the total variation was only 5.2%, confirming the low correlation between DMI and ADG in feedlot trials. The correlation between metabolizable energy intake (MEI) and MEI/ADG also was low, but 80% of the variation in MEI/ADG was accounted for by the variation in ADG. Intake patterns, regardless of dietary energy concentration, indicate that ADG increased with DMI only when ADG was below 1.5 kg/d. Therefore, current prediction equations apparently overpredict ME requirements for fast-growing feedlot steers. Optimal performance occurred at a ME concentration of 12.0 MJ/kg of DM rather than 12.5 MJ/kg of DM. Dressing percentage and carcass gain increased as dietary energy concentration and ADG increased, and current prediction equations should predict that of fast-growing feedlot steers equally well.