Bone remains of small vertebrate fossils provide valuable information for paleoenvironmental and paleoclimatic reconstructions. However, direct radiocarbon dating of small vertebrates remains challenging as the extraction of sufficient good quality collagen is required. The efficiency of eight collagen extraction protocols was tested on seven samples, representative of different ages and burial environments, including both macro and small vertebrate taxa. First, the samples were prescreened using attenuated total reflectance–Fourier transform infrared spectroscopy (ATR-FTIR) to quantify collagen content in archaeological bones, revealing that one should be discarded for 14C dating. Then, the quantity of protein extracted (yield) and collagen integrity were checked using conventional elemental analysis. The results show that one protocol was not able to accurately extract collagen from the samples. A soft HCl-based protocol seems more appropriate for the pretreatment of archaeological small mammal bones, whereas a harsher protocol might be more efficient to extract a higher amount of collagen from large mammals as well as amphibian bones. The influence of the tested protocols on carbon and nitrogen isotope values was also investigated. The results showed that isotopic variability, when existing, is related to the interindividual differences rather than the different protocols.