We have compared the use of new markers of bone turnover in the assessment and treatment of Paget disease and made observations on the mechanisms of bone resorption. Urine hydroxyproline (Hyp) as a bone resorption marker and serum alkaline phosphatase (ALP) as a bone formation marker have traditionally been used to biochemically assess and monitor treatment of Paget disease. Hyp and total ALP were compared with total urine pyridinoline (Pyd) and deoxypyridinoline (Dpd), free urine Pyd and Dpd, urine type I collagen N-terminal cross-linked telopeptide (NTX), type I collagen C-terminal propeptide (PICP), serum osteocalcin, and bone ALP in Paget patients treated with pamidronate. Patients were divided into three biochemical severity-based treatment groups by their fasting urine hydroxyprolline excretion (HypE) levels (Le., group 1, HypE < 5.0 mumol/l of glomerular filtrate [GF]; group 2, HypE of 5.0-9.9 mumol/l of GF; group 3, HypE > 10 mumol/l of GF). Group 1 received one 60 mg intravenous infusion of pamidronate, and groups 2 and 3 received four and six 60 mg infusions at weekly intervals, respectively. Fasting serum and morning urine specimens were taken before and at 2, 6, 13, and 26 weeks after starting treatment. Baseline Z scores were used to compare separation of patient results from normal, and the difference in Z scores from baseline to 13 weeks was used to compare response to treatment. Baseline discrimination and response to treatment at all disease activity levels was greatest for NTX and was poor for osteocalcin, PICP, and C-terminal cross-linked telopeptide of type I collagen (ICTP). The other markers showed good discrimination and response at medium and high levels of disease activity. NTX, total Pyd and Dpd, free Pyd and Dpd, and ICTP are all pyridinoline cross-link-based markers, but discrimination and response by NTX was generally much greater than for the others. Determination of the mechanism of the difference between NTX and other cross-link measures is necessary for appropriate use of the markers and may also lead to a better understanding of the bone resorption process. It has been proposed that the greater sensitivity and discrimination of NTX is because it is more bone-specific than the other cross-link markers with significant amounts of free Pyd and Dpd coming from nonbone sources. We propose another model where the proportion of peptide-bound cross-links such as NTX may be increased in high bone turnover states partly due to a rate-limiting step in their degradation to free cross-links. Conditions with high bone resorption rates would have high levels of NTX that would decline rapidly when resorption rates fall to a level where the capacity to degrade NTX matches the rate of production.