With the increasing demand for clinically useful biomarkers of bone turnover, a number of assays for the measurement of bone resorption markers have been developed. In the present study, automated (ACS: 180 DPD, Chiron Diagnostics, USA) and manual (DPD-ELISA, Pyrilinks-D, Metra Biosystems, USA) immunoassays for free DPD, and a manual immunoassay for the aminoterminal telopeptide of type I collagen (NTX, Osteomark, Ostex International, USA) were compared to the automated HPLC method for free DPD. Urine samples from a total of 538 healthy and diseased subjects aged 20 to 80 years were analyzed. The age and sex stratified reference ranges were essentially identical for the HPLC, ACS: 180 and the DPD-ELISA, but differed from the NTX assay. Individual values for free DPD as generated by HPLC and immunoassay techniques were highly correlated with each other, whereas correlations between assays measuring free and peptide-bound crosslink components were less pronounced. Precision of the automated techniques (HPLC and ACS: 180) was superior to that of the manual immunoassays. Disease-specific changes in crosslink excretion were similar for all assays and most pronounced in metastatic osteopathy, primary hyperparathyroidism and untreated Paget's disease of bone. We conclude that the automated assays for free DPD in urine, i.e. the HPLC and the ACS: 180 assay, show better analytical performance than the manual immunoassays studied. All techniques used in the present study appear to provide similar or identical clinical information. Therefore, the decision which assay to use largely depends on the laboratory set-up, the number of samples to be analysed, the turn-around time required, and the application for which the test should be used.