27 patients on hemodialysis (dialysate aluminium < 0.7 μmol/l for 2 years, and 2 μmol/l before) whose plasma Ca and PO<sub>4</sub> were adequately controlled for already 6 months by high doses of CaCO<sub>3</sub> alone (mean ± SD: 9 ± 5 g/day), were randomly divided into 2 groups, a control group (c group) which was kept on the same treatment, and a group in which CaCO<sub>3</sub> was reduced to 3 g/day but in which plasma Ca was kept normal due to 1α-OH-vitamin D<sub>3</sub> administration (1 μg/day at the beginning, 0.3 μg/day after 6 months; 1α group) whereas plasma phosphate was kept below 6.0 mg/dl because of A1(OH)<sub>3</sub> (2.7–5 g/day). Initially, the 2 groups were comparable as regards the plasma concentrations of total and ionized Ca, phosphate, alkaline phosphatases, medium and C-terminal parathyroid hormone (PTH) and aluminium, but the control group had lower plasma 25-OH-vitamin D (25-OHD.) After 6 months, the same difference in plasma 25-OHD was found with comparable plasma concentrations of total and ionized calcium as well as of medium and C-terminal PTH (beta error 1%). However, plasma concentration of phosphate and the plasma Ca phosphate product, as well as the plasma aluminium were higher in the 1α group whereas their PCO<sub>3</sub>H was lower. Although the alkaline phosphatase values were not significantly different between the 2 groups, they increased only in the control group because of 1 patient who developed a vitamin-D-deficient osteomalacia (plasma 25-OHD 3 ng/ml), which was subsequently cured by physiological doses of 25-OHD<sub>3</sub>. The incidence of transient hypercalcemia (15 vs. 21 episodes) and worsening of soft tissue calcifications (3 in each group) was the same in the 2 groups.