We compared methods of assessing the diet of the bearded vulture (Gypaetus barbatus) during the nestling period in the Pyrenees, northeast Spain. We determined diet from direct observations of food items delivered to the nest, recent prey remains present in the nest, remains collected in the nests after fledging, and remains collected in the ossuaries (bone-breaking sites). Data suggest that direct observation (food items delivered and recent prey remains present in the nest) is the only valid method of assessing the bearded vulture's diet accurately. Remains overestimated the presence of large mammals, such as cows (Bos taurus) and horses (Equus caballus), Suidae, and birds; delivered samples contained a higher proportion of small mammals, medium-sized mammals, micromammals, and reptiles. Ossuaries also differed from delivered samples because remains collected there overestimated large and medium-sized mammals. Concerning the skeletal parts, ossuaries, compared to all other methods, underestimated extremities and overestimated long bones, such as femur, humeri and tibiae, scapulae, vertebrae, and skulls. Remains samples, which overestimated scapulae, also differed from delivered and present samples. Our results suggest that bearded vultures favor extremities of prey (78% of the mammal remains, which make up 95% of the diet). The prevalence of small carcasses (almost 17%) suggests that vultures select small animals for food for the young. Because food quality may influence breeding success, future conservation projects based on the selective provision of food to breeding pairs should add to food stations meat remains and small carcasses consistent with our assessment of the birds' dietary needs.