Biomolecular archaeology is a rapidly expanding field which uses scientific techniques to investigate organic remains from archaeological sites. Proteins and DNA are the most biologically informative molecules that can be studied in ancient bone, although the analysis of ancient DNA (aDNA) has dominated the field for the past thirty years despite problems with sample contamination with extant DNA. Advances in proteomics techniques in the last two decades have resulted in sensitive instruments which can detect very low concentrations (picomoles) of proteins in complex mixtures, and which are suitable for the analysis of potentially damaged ancient proteins due to low-energy 'soft' ionisation techniques. The development of such techniques combined with the fact that proteins can provide biologically informative data while avoiding some of the problems facing aDNA analysis has caused a recent increase in interest in the study of ancient proteins. Despite this, little is still known about the ancient bone proteome and the potential longevity of useful proteins in ancient bone. This thesis uses conventional shotgun proteomics methodologies to investigate the temporal limits of protein survival and the potential utility of non-collagenous proteins (NCPs) in ancient bovine bone. Liquid chromatography tandem mass spectrometry (LC-MS/MS) was applied to a range of ancient bovine bone samples of different ages (10,000 to 1.5 million years) from different burial environments. The results presented here indicate that collagen survives well in ancient bone, being identified in the oldest sample tested here (1.5 million years old) and that NCPs are relatively long lived and survive in detectable quantities in bone samples up to 900,000 years old. Additionally it appears that some of these long-lived NCPs, such as fetuin-A, have enough sequence variation between species that they may be useful for phylogenetic analyses. This study has also shown that waste fractions from stable isotope and radiocarbon dating analyses of ancient bone, which are routine in archaeological science, contain useful NCPs and therefore may be a source of biological information that is currently being overlooked. One attractive potential application of ancient proteomics is as a screening method for the presence of aDNA in an ancient bone sample. Such an approach would require the identification of protein biomarkers which correlate with the presence or absence of aDNA in a sample. However, the attempts made in this thesis to assess the efficacy of proteomics as a screening method for the presence of aDNA by comparing mtDNA and proteome data from ancient bone failed to identify any suitable protein biomarkers. This suggests that proteomics may be useful in phylogenetic analyses of samples which are older than the proposed survival limits of aDNA but is not a useful screening method for aDNA in ancient bone. Additionally, proteins and DNA probably survive in ancient bone via different mechanisms; aDNA preservation may be more reliant on adsorption to hydroxyapatite crystals in bone.