The ability to authenticate the feed given to animals from the animal products has become a major challenge for scientists, monitoring bodies and commercial entities alike. This study compared two methods based on the use of the visible reflectance spectrum of the fat to discriminate pasture-fed (P) from stall concentrate-fed (S) lamb carcasses. A total of 307 (143 P and 164 S) Limousine lambs were used over 2 years. Pasture-fed lambs grazed a permanent pasture that was maintained at a leafy, green vegetative stage, and offered ad libitum; they received no supplementation at pasture. Body weight of P lambs when turning out to pasture and at slaughter averaged 9.2 (standard deviation (s.d.) 2.21) kg and 33.2 (s.d. 2.89) kg, respectively. S lambs were fed indoors on an ad libitum diet of commercial concentrate and hay until slaughter at a mean body weight of 33.7 (s.d. 3.62) kg. The reflectance spectrum of perirenal and subcutaneous caudal fat was measured at slaughter and at 24 h post mortem. Plasma carotenoid concentration was measured at slaughter. In method 1, the fat reflectance spectrum data were used at wavelengths between 450 and 510 nm to calculate an index quantifying light absorption by carotenoid pigments. In method 2, a multivariate analysis was performed over the full set of fat reflectance data at wavelengths between 400 and 700 nm. Method 2 yielded a higher proportion of correctly classified lambs compared with method 1 (P < 0.05 to 0.001), except for measurements made at 24 h post mortem on perirenal fat for S lambs. The proportion of lambs correctly classified using method 2 was 87.4% and 92.9% for measurements made on perirenal and caudal fat at slaughter, and 93.9% and 91.0% for measurements made on perirenal and caudal fat 24 h post mortem. Plasma carotenoid concentrations were higher in P lambs than in S lambs (P < 0.001), which led to correct classification of 90.7% of the lambs.