Pamidronate has been demonstrated to be an effective agent in the treatment of cancer-associated hypercalcaemia. The dose regime, however, remains controversial. In this study 16 patients with cancer-associated hypercalcaemia were given 30 mg pamidronate by intravenous infusion and 16 were given 90 mg also by infusion. Groups were well-matched in terms of tumour types, bone metastases, pre-treatment serum calcium and creatinine, fasting urinary calcium/creatinine ratio, nephrogenous cAMP and the renal tubular threshold for phosphate reabsorption (TmPO4). The calcium lowering effect was similar in both treatment groups with nadir at day 6 of mean (+/- SEM) 2.48 mmol/l (+/- 0.06) in the 30 mg group and at day 9 in the 90 mg group of 2.51 mmol/l (+/- 0.03) (P less than 0.01). 10 patients in the 30 mg group and 8 in the 90 mg group were normocalcaemic at this point. Similarly when those patients with more severe hypercalcaemia (greater than 3.30 mmol/l, n = 7 in each group) were analysed separately, no significant difference was evident between the two groups. Urinary calcium/creatinine ratios fell to a nadir at day 6 in both groups of 0.33 (+/- 0.05) (30 mg group) and 0.37 (+/- 0.10) (90 mg group) (P less than 0.01). Follow-up results after the initial 9 days showed the mean time to relapse to be 38 days (range 18-90) in the 30 mg group and 34 days (11-105) in the 90 mg group.(ABSTRACT TRUNCATED AT 250 WORDS)