BACKGROUND Stable isotope analysis is widely used to reconstruct diet, delineate trophic interactions, and determine energy pathways. Such ecological inferences are based on the idea that animals are, isotopically, what they eat but with a predictable difference between the isotopic ratio of a consumer and that of its diet, coined as the discrimination factor. Providing correct estimates of diet-consumer isotopic discrimination in controlled conditions is key for a robust application of the stable isotopes technique in the wild. METHODS Using a Finnigan Mat Delta Plus isotope-ratio mass spectrometer, we investigated isotopic discrimination of carbon and nitrogen isotope ratios (δ13 C and δ15 N values) in guard hairs of four Arctic predators; the wolf (n = 7), the wolverine (n = 2), the grizzly bear (n = 2), and the polar bear (n = 3). During a 3-month trial, carnivores were fed a mixed diet. The δ13 C and δ15 N values, and the mass (g) of diet items, were monitored weekly for each individual to determine their Total Diet Average ratios. RESULTS Diet-hair isotopic discrimination (Δx) varied according to species, ranging [1.88 ± 0.69‰: 3.2 ± 0.69‰] for δ13 C values, and [1.58 ± 0.17‰: 3.81 ± 0.22‰] for δ15 N values. Adult wolves Δ13 C average (2.03 ± 0.7‰) was lower than that of young wolves (2.60 ± 0.8‰) and any other species (combined average of 2.59 ± 0.28‰), except for the wolverine (2.12 ± 0.23‰). Wolves Δ15 N averages (juveniles: 3.51 ± 0.34‰, adults: 3.68 ± 0.28‰) were higher than those of any other species (combined average: 2.50 ± 0.58‰). CONCLUSIONS The discrimination factors for δ13 C and δ15 N values calculated in this study could be used in ecological studies dealing with free-ranging animals, with implications for non-invasive research approaches. As in other controlled discrimination studies, we recommend caution in applying our discrimination factors when the population structure is heterogeneous.