Few data are available on the ability of bone markers to predict the symptomatic response to bisphosphonate therapy in patients with painful bone metastases. We evaluated the levels of bone markers in patients with bone metastases receiving pamidronate and determined the corresponding analgesic response. Forty-two patients were administered two two-week cycles of intravenous pamidronate 60 mg/week with a three-week interval in between. Serum levels of bone formation, resorption and other bone-associated markers (osteoprotegerin, osteopontin and calcium) were measured. Levels of two urinary markers were also measured and the intensity of pain and analgesic drug consumption evaluated. A mixed effects linear modelling approach was adopted to account for possible correlation among marker levels and time on study or analgesic response. We created an indicator variable that classified the patients' analgesic response as 'improved/stationary' or 'worsened' determined by patient reported intensity of pain and analgesic drug consumption. Eighteen patients 'worsened' and 24 were 'improved/stationary'. The results of the mixed effects models for testing the association between marker levels and time on study or analgesic response showed: i) the changes in marker levels over time did not significantly differ between the two groups; ii) the overall test for time on study was not statistically significant for C-terminal telopeptide of type I collagen (ICTP), osteoprotegerin and osteopontin; iii) in contrast, ICTP and osteoprotegerin were significantly associated with analgesic response. Biochemical markers of bone turnover, in particular ICTP and osteoprotegerin seem promising for predicting and objectively assessing the analgesic response to pamidronate treatment.