The influence of aluminium in dialysate on the effects of 1α(OH)3 on hemodialyzed hypocalcemic patients with end-stage renal failure, was studied during a 24- to 42-month period. 51 hypocalcemic patients were divided into two groups; group 1 consisted of 28 patients who were dialyzed using dialysate prepared from reverse osmosed water; the 23 patients in group 2 used dialysate prepared from softened water. Aluminium concentration in the dialysate used for group 1 was less than the detectable limit (10 µg/l) in twelve times determinations, while that for group 2 was 23.1 ± 9.2 µg/l (mean ± SD, n = 14). By the administration of 1α(OH)D3, the serum concentration of calcium was increased, and that of iPTH and alkaline phosphatase activity was decreased in both groups. Subperiosteal resorption of the finger bone, evaluated by Jensen’s criteria, was significantly improved in group 1, while there was no improvement in group 2. Serum aluminium concentration in the patients of group 1 and group 2 were 46.6 ± 6.3 and 84.7 ± 13.9 µg/l, respectively, and the concentration of the latter was significantly higher than that of the former (p < 0.01). It was also shown that there is a positive correlation between the extent of subperiosteal resorption and the concentration of aluminium in serum. Serum aluminium concentration and bone aluminium content were increased according to the duration of hemodialysis in the patients who were dialyzed using dialysate from softened water, while there was no correlation between the duration of hemodialysis and serum aluminium concentration for the patients of group 1. After changing the supply of water for dialysate, prepared from softened water to reverse osmosed water, serum aluminium concentration significantly decreased. These results are consistent with, but do not prove, the hypothesis that aluminium plays a pathological role in renal oste