Type I collagen is the major bone protein. Little is known quantitatively about human bone collagen synthesis in vivo, despite its importance for the understanding of bone formation and turnover. Our aim was to develop a method that could be used for the physiological and pathophysiological investigation of human bone collagen synthesis. We have carried out preliminary studies in patients undergoing hip replacement and in pigs to validate the use of the flooding dose method using (13)C- or (15)N-labelled proline and we have now refined our techniques to allow them to be used in a normal clinical or physiological setting. The results show that the application of a flooding dose causes bone free-proline labelling to equilibrate with that of blood in pigs and human beings, so that only 150 mg of bone will provide enough sample to prepare and measure the labelling of three fractions of bone collagen (dissolved in NaCl, acetic acid and pepsin/acetic acid) which have the same relative labelling (1.0:0.43:0.1) as measured by GC-combustion-isotope ratio MS. The rates of incorporation were substantially faster than in skeletal muscle samples taken at the same time. The results suggest that different fractions of human bone collagen turnover at markedly higher rates than had been previously considered. This approach should allow us to discover how growth and development, food, activity and drugs affect bone collagen turnover and to measure the effects on it of ageing and bone disease.