Preclinical studies have shown that the vitamin D analogue EB 1089 has significantly less calcaemic activity than its parent compound 1,25-dihydroxyvitamin D (1,25(OH)2D3) and significant anti-tumour activity. This phase I trial was designed to evaluate the calcaemic effect of the drug in patients with advanced cancer. EB 1089 was given to 36 patients with advanced breast and colorectal cancer in doses of between 0.15 and 17.0 microg m(-2) day(-1). Serial serum and urine calcium, urine creatinine and serum parathyroid hormone (PTH) were monitored. Hypercalcaemia was seen in all patients receiving 17.0 microg m(-2) day(-1). Hypercalcaemia attributable to EB 1089 was reversible by discontinuing or reducing EB 1089 therapy. During the first 5 days of treatment, urine calcium (P = 0.0001) and serum-corrected calcium (P = 0.027) were related to EB 1089 dose, whereas serum parathyroid hormone (P = 0.0001) showed an inverse relationship. Twenty-one patients received compassionate treatment for between 10 and 234 days. No complete or partial responses were seen. Six patients on treatment for more than 90 days showed stabilization of disease. EB 1089 was well tolerated and adverse events considered to be caused by EB 1089 were limited to dose-dependent effects on calcium metabolism. The dose estimated to be tolerable for most patients from this study is around 7 microg m(-2) day(1). These data support previous work that has demonstrated EB 1089 to be significantly less calcaemic than 1,25-dihydroxyvitamin D3.