The study of human diet in the past investigates the presence and post-depositional survival of indicators to identify human interactions with animals and plants. When present, their recognition requires the use of high-resolution images to distinguish their provenance. Nevertheless, inconspicuous surface modifications, such as human tooth scores, are difficult to identify even through high quality images. The use of laser scanning confocal microscopy offers quality images, which when combined with morphometric values, provide means of differentiating between different types of anthropogenic bone surface modifications. This paper presents a neotaphonomic experimental study testing whether measurement types and magnification level affect the accuracy of established identifications methods for grouping different categories of human modifications related to the consumption of small vertebrates. Through the multivariate statistical analyses of repeated measures, this paper proposes the best combination of parameters to identify human tooth scores in small vertebrates using laser scanning confocal microscopy. The results have implications for the study of human consumption of small prey, including rabbits, rodents, birds and reptiles, a key piece in the research of human occupation patterns and behaviour during the Pleistocene in Eurasia and Island Southeast Asia.