The scope of this article is to propose an innovative method based on Near Infrared Hyperspectral Chemical Imaging (NIR-HCI) to rapidly and non-destructively evaluate the relative degree of collagen preservation in bones recovered from archaeological contexts. This preliminary study has allowed the evaluation of the potential of the method using bone samples from the Early Upper Palaeolithic, Mesolithic and Neolithic periods at the site of Trou AlWesse in Belgium. NIR-HCI, combined with chemometric tools, has identified specific spectral bands characteristic of collagen. A chemometric model has been built using Partial Least Squares Discriminant Analysis (PLS-DA) to identify bones with and without collagen. This enables the evaluation of the degree of collagen preservation and homogeneity in bones within and between different strata, which has direct implications for archaeological applications (e.g., taphonomic analyses, assemblage integrity) and sample selection for subsequent analyses requiring collagen. Two archaeological applications are presented: comparison between sub-layers in an Early Upper Palaeolithic unit, and evaluation of the range of variability in collagen preservation within a single Holocene stratum.