In the western world cancer is the second most important cause of death, after heart disease, accounting for 20-25% of all mortalities, and is today probably one of the most feared diseases. Approximately one third of all cancer patients will develop skeletal metastases (secondary bone cancer). Early diagnosis and effective monitoring during treatment is therefore essential in terms of making any impact on survival rates and developing new cancer therapies. Unfortunately, the current methods for diagnosing and measuring bone metastases, such as bone scans and urinary hydroxyproline determinations, lack sensitivity and specificity. The urinary pyridinium crosslinks, pyridinoline (PYD) and deoxypyridinoline (DPYD), have recently been identified as sensitive and specific markers of bone breakdown. However the analysis of the pyridinium crosslinks using high performance liquid chromatography (HPLC) has proved far from ideal for routine clinical assessment. The results from studies to critically evaluate this method are presented and particular problems encountered when the crosslinks are extracted from the urine samples are discussed. The tedious, time consuming and cumbersome sample preparation procedure are also shown to adversely effect the robustness and reproducibility of the method. The recent introduction of an immunoassay method potentially overcomes many of the inherent problems with the HPLC analysis. This enzyme linked immunoasorbant assay (ELISA) is evaluated and found to compare favourably with the HPLC method, offering several distinct advantages. The method is quick, simple, robust, demonstrates good accuracy and precision, is less prone to interferences and can be easily introduced into clinical laboratories on a routine basis. In addition it also minimises sample preparation time. However, there is still a requirement for alternative and better biochemical markers to measure bone breakdown. It is well known that bone is an active, living tissue and that bone metabolism and remodelling are tightly coupled processes, where the rate of bone formation equals the rate of bone resorption in healthy bone. When an imbalance occurs, this leads to unhealthy bone and ultimately a clinical disease of the skeleton. Some trace metals, e.g lead, accumulate in the bone and since the development of bone metastases results in extensive bone breakdown, the subsequent release of these metals into the blood and urine may potentially serve as markers. In this work inductively coupled plasma-mass spectrometry (ICP-MS) has been used and methods developed to determine such metals in clinical matrices. The development of a simple dilution method is described for use in preliminary trials to measure the blood lead levels and other trace metal profiles, in patients with bone metastases. The blood lead results attained agree closely with a certified reference material, and the method is shown to remain under analytical control. The trial results are presented and discussed with reference to further and more detailed investigations. The selection criteria for other suitable elements such as AI, Ba, Cd, Ce, Pb, Sr, and Zr in blood and urine, along with an assessment of the analytical and clinical praticalities of the methodology which must be considered for subsequent full clinical trials is also discussed following a critical evaluation. Finally the results obtained in a extended clinical trial are presented. The crosslink levels (serving as the reference marker), measured by ELISA, were compared with the trace metal levels (Cd, Pb and Sr) in blood and urine samples, measured by ICP-MS, in order to assess their diagnostic potential and effectiveness in monitoring treatment. The blood lead levels were found to offer the greatest potential, correlating well with the DPYD values in the majority of cases. The blood strontium levels also showed some promise. However the blood cadmium and the urinary trace metal levels proved less suitable. The results attained in this feasibility study support a more detailed clinical investigation, on a much larger scale, and over a longer period of time. The need to incorporate a full statistical evaluation of all factors that can influence the final results is highlighted.