Three fattening systems were evaluated from weaning to slaughter in order to find alternatives to grain feeding in young bulls, and to test the reliability of carcass subcutaneous fat colour to discriminate among them. After weaning (224 kg), one group of animals was fed concentrates and straw until they reached the target slaughter weight (450 kg; Feedlot), another group grazed rotationally on lucerne supplemented with 1.8 kg DM/day barley until slaughter (LUC), and the third group had the same management as LUC animals for 3 months (period 1) and thereafter was finished on concentrates and straw until slaughter (period 2; LUC + Feedlot). Animals were weighed weekly and sampled monthly for serum IGF-I and leptin, and plasma non-esterified fatty acids and carotenoid pigment concentration analyses. Carcass characteristics and subcutaneous fat colour were recorded after slaughter. In period 1, Feedlot animals had slightly greater weight gains than their grazing counterparts (P < 0.10), and at the end of period 1 they had 66% greater IGF-I and 35% greater leptin concentration (P < 0.01). Plasma carotenoid pigments were undetectable in Feedlot animals, but increased during grazing in LUC and LUC + Feedlot treatments. In period 2, weight gains were lowest for LUC, intermediate for Feedlot and greatest for LUC + Feedlot animals (P < 0.001), conditioning the time taken to reach slaughter weight (73, 58 and 47 days, respectively; P < 0.05). Leptin and IGF-I concentrations increased in all management systems during period 2. Plasma carotenoid pigment concentration reached its maximum at the end of period 2 in LUC animals, but it decreased sharply in LUC + Feedlot animals in this period. Management did not affect carcass traits except for subcutaneous fat colour. Yellowness, Chroma (C*) and the value of the integral of the translated reflectance spectrum (SUM), estimator of carotenoid pigment content in fat, were higher in LUC than in LUC + Feedlot and Feedlot animals (P < 0.001). Two logistic regressions were obtained to discriminate carcasses from LUC treatment: P (LUC) = (1 + e18.8–5.6 × lightness–36.9 × redness + 0.3 × SUM + 29.8 × C*)−1 and LUC + Feedlot treatment: P (LUC + Feedlot)=(1 + e833.7–11.8 × lightness + 4.7 × redness + 0.2 × SUM−2.5 × C*)−1. The economic margin, calculated as income achieved minus costs, was greatest for LUC, intermediate for LUC + Feedlot and lowest for Feedlot treatment. Therefore, grazing lucerne supplemented with barley was an interesting alternative for fattening young bulls in these conditions, producing carcasses of similar quality, which could be accurately traced by measuring subcutaneous fat colour.