To find out if the concentration of parathyroid hormone-related protein (PTHrP) predicts the response of tumour-inducing hypercalcaemia (TIH) to pamidronate, we studied 44 patients. Pretreatment measurements of serum PTHrP, calcium and phosphate, nephrogenous cyclic AMP, tubular threshold for calcium and phosphate (TmP), and the presence of bone metastases were correlated with response to pamidronate. Response was considered good (normal calcium concentration corrected for albumin [CCa] for >14 days), or poor (failure of CCa to fall, or a rise above normal ≤ 14 days). PTHrP correlated significantly with response (good vs poor, p=0·02). Undetectable PTHrP (<2 pmol/L) was associated with a good response in all seven treatments, PTHrP in the range 2-12 pmol/L was associated with good response in 10 of 14 treatments, while PTHrP ≥12 pmol/L was associated with a poor response in all 11 treatments. Tubular threshold for calcium correlated with the fall in CCa by day 6 after treatment (p=0·02). Urinary clearance estimations in poor responders suggested that there was an incomplete reversal of the renal tubular component of hypercalcaemia. Serum PTHrP correlates with response to pamidronate in the treatment of TIH; which may be associated with a renal tubular mechanism not significantly affected by currently available treatment. Drugs that inhibit tubular reabsorption of calcium or PTHrP secretion may help in patients who do not respond to pamidronate.