The purpose of this study is to evaluate the place of intravenous 1α-hydroxyvitamin D3 (1α-OH-D3) in the prevention of radiologically obvious hyperparathyroidism (HPT) in patients on maintenance dialysis while excluding aluminium phosphate binder and using a dialysate calcium concentration of 1.62 mmol which keeps the intradialytic calcium balance neutral. Therefore, 47 patients without subperiosteal resorption and previously treated by oral CaCO3 and if necessary Mg(OH)2 as phosphate binder while their dialysate calcium had a Ca level of 1.62 and a Mg level of 0.2 mmol/l were randomized into a control group of 24 who were maintained on the same treatment and an experimental group of 23. This group discontinued CaCO3 and received intravenous 1α-OH-D3 after each dialysis at increasing doses up to 4 μg and increased Mg(OH)2 as their sole phosphate binder. When plasma Ca increased above 2.7 mmol/l, the dose of 1α-OH-D3 was decreased. When plasma P04 increased above 2 mmol/l, the dose of Mg(OH)2 was increased to the highest dose not inducing diarrhea, hypermagnesemia ( < 2 mmol/l) or hyperkalemia ( < 6 mmol/l). In case of persistent hyperphosphatemia, the dose of 1α-OH-D3 was decreased. Since mean plasma alcaline phosphatase was normal, HPT was monitored on the plasma concentration of 1-84 PTH for which a previous histological study showed that frank osteitis fíbrosa was present only when they were above 70 pg/ml, i.e. (about twice the upper limit of the normal value). Before the study, plasma PTH was below this limit in 16 patients of the CaCO3 group and in 14 patients of the 1α-OH-D3 group. After 6 months, they remained below this limit in all patients except 2 of each group. Plasma PTH was initially above 70 pg/ml in 8 of the CaCO3 and did not change significantly throughout the study, 2 patients having at 6 months a PTH level below 70 pg/ml. In contrast with intravenous 1α-OH-D3, all the 9 patients with initial frank HPT decreased their PTH levels after 2 months, the levels being below 70 pg/ml in 6 cases. However, because of hypercalcemia and/or of hyperphosphatemia in spite of a highest tolerable dose of Mg(OH)2, 1α-OH-D3 doses had to be decreased down to 0.4 μg per dialysis at the 6th month so that at 6 months 6 of 9 patients had their PTH levels above 70 pg/ml, a number comparable to that of patients treated with CaCO3 (6 of 8). Conclusion: (l) CaCO3 and intravenous 1α-OH-D3 are comparably capable of preventing HPT in two thirds of the dialyzed patients; (2) only intravenous 1α-OH-D3 is able to correct an established frank HPT but this correction is only transient when Mg(OH)2 is the sole phosphate binder. As a matter of fact, the highest tolerable doses of Mg(OH)2 cannot prevent the hyperphosphatemia worsening induced by intravenous 1α-OH-D3, which leads to a decrease of the dose of 1α-OH