A new immunoassay using an ELISA approach for measuring urinary excretion of cross-linked N-telopeptides of type 1 collagen was evaluated as a specific measure of bone resorption. The assay was applied to 65 early postmenopausal women who participated in a placebo-controlled trial of the aminobisphosphonate, alendronate sodium. Eight blood and urine samples were collected over a 9 month interval. Baseline cross-linked peptide excretion varied from 26 to 216 pmol BCE (bone collagen/mumol Cr. Within-subject variability (CV) for cross-linked peptide excretion was 20.2% over the 9 months in placebo-treated subjects, substantially less than that observed for other biochemical markers of bone resorption: 45, 53, and 63% for fasting urinary calcium and hydroxyproline and 24 h urinary lysylpyridinoline (HPLC assay), respectively. Baseline cross-linked peptide excretion correlated significantly (p < 0.001) with baseline total urine lysylpyridinoline and serum osteocalcin, but not with the other biochemical markers. Initial peptide excretion also correlated inversely with lumbar spine bone mineral density at entry (r = -0.26, p < 0.05). Treatment for 6 weeks with alendronate produced a dose-dependent suppression of cross-linked peptide excretion (0 +/- 8, 29 +/- 6, 56 +/- 5, and 64 +/- 3% for 0, 5, 20, and 40 mg, respectively, p < 0.01 versus placebo for treatment effect), with a return toward pretreatment values during follow-up. Measurement of the urinary cross-linked N-telopeptides of type I collagen by this new ELISA approach appears promising as a simple and reliable method to assess overall bone resorption.(ABSTRACT TRUNCATED AT 250 WORDS)