Twenty-four 3-month-old St. Croix wethers (14.5 ± 0.53 kg initial live weight) were used to determine potential for enhancing efficiency of feed utilization by feeding a growing diet marginal in protein (83 days) to elevate fat deposition and partially delay protein accretion until a subsequent finishing phase (65 days; 80% concentrate, 20% crude protein diet). Endophyte-free fescue hay (14.4% crude protein) was consumed ad libitum during growing (83 days), with 1.25% body weight of maize plus a mixture of high-protein feedstuffs (82% crude protein) (LM-P), 2.5% body weight of maize plus high-protein feedstuffs (HM-P), or 2.5% body weight of maize (HM). Growing phase dry matter intake was similar among treatments (923, 978, and 900 g/day; SE 64.1) and live weight gain was greater (P<0.05) for HM-P vs LM-P and HM (139, 188, and 148 g/day for LM-P, HM-P, and H M , respectively; SE 8.9). Empty body protein mass after growing ranked (P<0.05) HM < LM-P < HM-P (4.17, 4.31, and 4.10 kg; SE 0.011), and fat mass was greater (P<0.05) for HM-P vs LM-P and HM (6.07, 7.01, and 5.66 kg for LM-P, HM-P, and HM, respectively; SE 0.188). Dry matter intake and live weight gain during, and body composition after, finishing were similar among treatments, although much of the difference in fat mass after growing between HM-P and other treatments was maintained through finishing (13.6, 14.2, and 13.6 kg for LM-P, HM-P, and HM, respectively; SE 0.80).