Radiocarbon dating of bones can be very useful in archaeological contexts, especially when dealing with funerary deposits lacking material culture, e.g. pottery vessels. 14C measurements of bone samples are usually performed on the extracted collagen residue. The content and the quality of collagen can vary significantly, mainly depending on bone preservation and diagenesis. Generally speaking, environmental conditions such as low pH level of soils, high temperatures, and percolating groundwaters, typical of arid and tropical zones, can affect the preservation of collagen; at the same time, bones recovered in such environments are more likely to be contaminated with carbon from the surrounding environment. Possible contamination of samples can also occur in temperate zones. While low collagen content is a condition we cannot overcome, we can use several chemical and elemental indicators in order to assess collagen quality. Among these, the C/N atomic ratio is considered a good parameter for detecting low-quality collagen and possibly contaminated samples. In a combustion and graphitization setup like that installed at INFN-LABEC, Florence, measurement can be easily performed using an elemental analyzer when combusting the sample prior to graphitization, thus requiring no extra effort (or extra amount of sample) during the preparation procedure. Bone samples recently 14C dated at INFN-LABEC have confirmed that the measurement of C/N atomic ratios can give some indications of the collagen quality. The bone material was collected from 3 necropoles of the Bronze Age period in Cyprus (Erimi-Laonin tou Porakou, Lophou-Kolaouzou, and Erimi-Kafkalla&Pitharka, along the Kouris Valley), an area characterized by environmental conditions that do not favor bone preservation. Samples were treated to extract collagen and measured by accelerator mass spectrometry (AMS). 14C results have been compared with the archaeological evidence, showing some relationship between measured C/N atomic ratios and collagen quality. In particular, when grouping the measured samples according to their C/N ratio, the agreement between 14C dates and archaeological evidence is good or inconsistent when the C/N ratio clearly falls inside or outside the “recommended” range, respectively, with a still reasonable agreement also when it is slightly above the upper limit of that range.